I’m on a quest for the holy grail of automated trading. Questions? Check the FAQs or feel free to reach out directly: AutomatedTradingStrategies@protonmail.com.

For Links To All Strategies Click Here

As you all know, I’ve been testing a weekly model for alerts. The current model selects strong candidates based on historical performance and current events/anomalies.

An anomaly, in the way I use it here, is a statistically recurring deviation from what random market behavior would predict. These aren’t one-off surprises—they’re patterns that keep showing up in the forward test data when they theoretically shouldn’t if markets were perfectly efficient.

I track several categories:

• Time-of-day anomalies are intraday patterns baked into market microstructure. The 11 AM Reversal is the flagship example—a measurable performance cliff that occurs when the European session overlap ends and US-only flow takes over. These tend to be the most persistent because they’re driven by structural mechanics rather than sentiment.

• Calendar anomalies are patterns tied to specific days-of-week or days-of-month. Week-2 Strength and Mid-Month Strength fall here. These likely reflect institutional rebalancing flows, options expiration cycles, and systematic fund activity that clusters around predictable dates.

• Event-driven anomalies are behavioral signatures the market leaves around scheduled economic releases—NFP, CPI, FOMC. I’m not predicting the number. I’m trading the shape of the market’s reaction: positioning ahead, whipsaw during, directional resolution after.

Every anomaly is derived empirically from my own Incubator and Forward Test data. They’re filtered through strategies I actually run.

One important caveat: not all anomalies are created equal. Some patterns are structural and rock-solid. Others are regime-dependent and shift year to year or even week to week. The alerts below are listed in order of persistence.

This Week’s Active Anomalies

These are the alerts for February 8–13, 2026.

Read more

I’m on a quest for the holy grail of automated trading. Questions? Check the FAQs or feel free to reach out directly: AutomatedTradingStrategies@protonmail.com.

For Links To All Strategies Click Here

Housekeeping: We have a lot of new subscribers this month, so welcome! If you have any questions about the process, please feel free to reach out via email or Substack Chat. It can be a bit overwhelming at first.

Okay, let’s get into it…

Every now and then I like to step back to see how far we’ve come and where we’re going. With so many new subscribers, this felt like the right time for an overview.

I can’t remember any quarter with this much going on. We started with HMM theory, then spent September and October reviewing strategy performance, with the last Mudder report going out in late October.

Currently, I am in the process of creating the Q1 2026 Forward Test. That involves looking at performance metrics over the last quarter as well as the full year. It also involves auditing strategies in the Incubator to determine what makes the cut for Q1 of 2026.

As you can see, we’re up $277K. The portfolio hit peak equity at $433K and is now in drawdown, but we still have two big weeks left in the quarter (FOMC, second week of month, last month of the year).

I’ll provide a comprehensive update on the Forward Test & Incubator later this week for paid subscribers. I think you’ll find the analysis highly valuable.

What Else Did We Do This Quarter?

I had some fun with the Ninjatrader Arena Cup Tournament. You can read more about my experience here:

These events represent a great way to test out your favorite trading strategy. I look forward to participating in the next one.

We also looked at a few automated trading strategies meant specifically to help new traders understand both where and when to trade by following “The Dog” and “The Cat.” These strategies visualize the best way to think about trade placement.

In Strategy 106, we talked about my process for creating a barbell strategy.

We aIso looked at VPS providers. We examined which features matter most in your trading setup. I’ll share a quick update on this today. Full project details here:

I also shared my favorite trade set-ups for 2026, which is all about energy.

Read the full post to learn about how AI data centers and tariffs are reshaping the entire commodities landscape, especially over the next year.

Automated Portfolio Managers: What I’m Most Excited For in 2026

The thing I’m most excited for in 2026 are Automated Portfolio Managers (APMs). When I’m running 20+ strategies under an APM:

• Some are in drawdown (expected)

• Some are thriving (expected)

• Some are being retired (expected)

• New ones are being promoted from incubation (expected)

When I’m betting on a single strategy, every drawdown is a crisis. Is it broken? Should I stop it? Am I about to give back all my gains?

If you follow this newsletter you know that over the last year, I’ve been obsessing on the subject of edge—which is constantly being chased by most strategies. Strategies are either in, out, or transitioning from degrees of edge, which expires completely for some strategies and marginally for others. This isn’t scalable. However, a system that can draw from different strategies depending on APM logic—HMM, contrarian, equity curve—just might be.

A single strategy is a hypothesis about markets. An APM is a hypothesis about how to continuously generate and validate hypotheses about markets.

The APM doesn’t need any single strategy to be immortal. It only needs the system to continuously identify and allocate to current edge. As I discussed in the post The Holy Grail Is A Process Not A Strategy, the edge isn’t in any individual strategy. The edge is in the process. APMs automate this process.

During the last Live Test two years ago, I had a difficult time with this “PM” role. As much as I want to believe that automated trading takes the emotion out of trading, in reality it only takes away my ability to act on that emotion. APMs create a secondary cushion of protection around the selection process. So you’re not only relying on the logic of the strategy, but the logic of the automated portfolio manager.

How are the APMs performing so far?

It’s been a rough start, but there is one APM that’s showing real promise. It made $63K last week on 741 trades. One PM made $5K over a three day period, and the other two performed poorly. The APMs are programmed to learn and tighten filters over time so performance should improve.

We’ve covered where we’ve been, now let me provide you with a quick EOY update on where some of these projects are today.

Read more

I’m on a quest for the holy grail of automated trading. Questions? Check the FAQs or feel free to reach out directly: AutomatedTradingStrategies@protonmail.com.

For Links To All Strategies Click Here

Housekeeping: I received the most charming email the other day. I was struck by the eloquence of word and insight. I thought, “This person should have their own Substack”. As I’ve said in the past, I’m not only a writer on Substack, but an investor. If you haven’t done so already, consider bringing your experience to Substack. My trading has improved dramatically since starting this newsletter and I’ve met some truly amazing people along the way. If you have any questions about the process, please feel free to reach out to me.

Okay, let’s get into it…

Executive Summary

Four weeks into Q4, the portfolio is up $74K on its own. But when I applied Hidden Markov Model (HMM) state-based selection to the trade data, it dramatically outperforms traditional day-of-week and time-of-day filtering. The most aggressive HMM approach (Option C) delivered $454K in net profits with position sizing, and a 72% daily win rate. I’m transitioning both the Forward Test and Alert System to this HMM-based approach.

Q4 Forward Test: Week 4 Performance Overview

Here’s a look at the daily P&L for the last four weeks:

Read more

I’m on a quest for the holy grail of automated trading. Questions? Check the FAQs or feel free to reach out directly: AutomatedTradingStrategies@protonmail.com.

For Links To All Strategies Click Here

Q4 Forward Test: Week 3

Well, this is the start of Week 4 of the Q4 Forward Test. Before talking about what to expect in Week 4, let’s discuss what happened in Week 3.

Three weeks into Q4 and we’re up $45K.

Read more

I’m on a quest for the holy grail of automated trading. Questions? Check the FAQs or feel free to reach out directly: AutomatedTradingStrategies@protonmail.com.

For Links To All Strategies Click Here

The Q3 2025 Forward Test

As a quick reminder, the Q3 Forward Test contains ~20 ATS trading strategies across 16 unique strategy types deployed on various instruments and timeframes.

Instruments Traded

• NQ (Nasdaq futures)

• MNQ (Micro Nasdaq)

• GC (Gold futures)

• ES (E-mini S&P 500)

• ZN (10-Year Treasury Note)

• ZB (30-Year Treasury Bond)

Data Series / Bar Frequency

• Minute-based: 22 strategies (ranging from 5-minute to 90-minute)

• Tick-based: 2 strategies (1,400 tick)

Account Structure

• All strategies run on simulated accounts

• Account naming convention: SimAccountStrategyNumberStrategyNumberStrategyNumberInstrumentInstrumentInstrumentTimeframeTimeframeTimeframe

• Some accounts run multiple instruments or timeframes

At the end of each quarter, I will remove poor performers and replace with standouts from the Incubator.

The Incubator is running variations of the first 100 strategies. In Q4, I’ll add embedded strategies.

The hope is to obtain as much actual data as possible for ML pattern recognition analysis by year’s end.

More Data

It’s all about the data this year. More data and higher-quality data improve performance. I’m also using the data to make revelations about trading a portfolio of strategies—from the selection process to capital allocation, every decision has an expectation that is constantly challenged with reality. Better data not only informs a clearer version of that reality, but helps to focus the lens where needed.

For example, we are not hunting volatility—we are hunting one-directional volatility. Most automated strategies struggle with change-point detection. Mean reversion is often used as a proxy, but it can be like threading a needle in trending markets. I’ve also learned that the best-performing days fall on days when nothing happens, which means the best pattern for automated trading strategies is a one-directional trend that occurs on days with no big announcements or events. While FOMC/CPI days have great potential, they are not as consistent.

Now let’s get into Q3 Forward Test performance and what the data tells us about taking Q4 to the next level. At the end, I’ll share the tentative Q4 Forward Test strategies and instruments with you.

Q3 Forward Test Performance

A few weeks ago I shared the Q3 Forward Test results with you.

At that time, the portfolio was up $70,165, but we were in a pretty severe drawdown. Here’s an overview of daily performance:

Here’s where we stand after September 17th:

While the 17th failed to impress, the portfolio still made a nice rebound from its drawdown low, going from $70,165 to $154,427. The largest drawdown was $90K, but the account never went lower than -$2,615.

Q3 End of Quarter Stats

• Final Portfolio Value: $154,427

• Lowest Equity Point: –$2,615 (portfolio was never lower than this)

• Maximum Drawdown: $89,606; 50% from a peak of $177,091 down to $87,485

• MAR (Return ÷ MaxDD): 1.72; 1.72 means it made $1.72 in net profit for every $1.00 of drawdown endured.

• Aggressive MAR (Return ÷ Lowest Equity Point): 59x. So the portfolio earned 59 dollars for every 1 dollar it dipped below breakeven at its lowest point.

• Sharpe (normalized to $100k): 1.21

• Sortino (normalized to $100k): 1.63

• Annualized return (on $100k proxy): 182%

• The best hours to enable trades were:

  • 4:00 AM ET: +$69,472 ($542.75 avg per trade)

  • 10:00 AM ET: +$46,706 ($491.64 avg)

  • 12:00 PM ET: +$34,673 ($412.78 avg)

• Day-of-week performance:

  • Monday/Tuesday dominance: +$177,553 combined profit

  • Sunday/Wednesday weakness: –$61,352 combined loss

• Other observations:

  • FOMC whipsaw pattern: Poor day-of performance, explosive next-day recovery (a new pattern).

  • High-volume days (>50 trades): Only 33% win rate, but positive expectancy.

  • Gold futures (GC): Massive volatility (–$109K in GC 12-25) requires position sizing adjustment.

Read more

We're not just developing strategies—we're on a quest for the holy grail of automated trading. Questions? Check the FAQs or feel free to reach out directly: AutomatedTradingStrategies@protonmail.com.

For Links To All Strategies Click Here

Housekeeping:

First, a few weeks ago I sent out an email informing you that someone was impersonating me. I have since been informed by Substack support that the profile has been deleted. Here’s the full response from Substack:

Hi Celan,

Tex from Substack Trust & Safety here. Thanks for reaching out about this. Reports like this one help us keep the platform safe for the entire Substack community.

It looks like we removed this impersonator over the weekend.

The fastest way to get impersonators removed is by having you or your subscribers follow these steps to report a profile through our web/app reporting flow. We're constantly improving auto-detection of those attempting to impersonate bestsellers but they sometimes get through.

Please don't hesitate to reach out if you have any further questions or concerns.

Best,

Tex @ Substack

So if you experience anything odd, contact Substack support (or me).

This is something that is only going to get worse with AI. Most recently, Sam Altman (the CEO of OpenAI) issued a warning to all banks regarding voice impersonation, so be sure to protect yourself.

Second, I’m so glad to be back! As I said before I left for vacation in August, all subscribers should have received one month as a comp. If you did not, please let me know.

As I make my way through your emails, I can’t tell you how thrilling it is for me to see your progress. I know how much the hunt means to me and it’s a blessing to pay some of that forward.

Third, I know some of you are putting off setting up a forward test of your own. I see you. One of my favorite pastimes is clicking “Watch Later” for no sugar dessert recipes on Youtube. Your trading system is becoming the exercise bike of your financial life.

Here's the thing: I spent TWO YEARS avoiding my first forward test (just 10 strategies). You don't need to purchase a subscription to start. Use the two free strategies (strategies 1 & 5) and just tweak them. Create variations. Break them. Fix them. The point isn't to find the perfect strategy—it's to start the process.

ATS strategies allow you to leap frog a step or two, but the holy grail is in the process and that process begins with the forward test. The good news is that what we’re hunting is automated. Once found, it’s an investment—however fleeting— in time.

Last but not least, many of you have already started your own forward test. The following is a framework for how to find redeeming value in any portfolio. Due to the number of trades, it also provides generalized information regarding systematic performance over Q3 that you may find translates to your own forward test. I’ve found that my portfolio tends to perform the same as many of the quant hedge funds I follow, which suggests that there’s some meta element which is a generalized algorithmic issue rather than a strategic one. I suspect it has everything to do with timing, but when I went looking for clues in the performance data, the culprit had more to do with volatility, directional volatility in particular.

Ultimately, the goal is to use this information to create a better portfolio for Q4, which starts in a few weeks.

With that said, let’s jump into the results of the Q3 2025 Forward Test, which is currently up $70K since June 19th.

The Q3 2025 Forward Test

The latest ATS Forward Test (Q3 2025) spans 2,791 trades across ES, NQ, GC, ZB/ZN, and multiple time-frames. The headline is simple: The portfolio made $70,165. Here’s an overview of daily performance:

Notice, the portfolio never dropped below $0. That has everything to do with start date, which was on a Monday. This was intentional.

Here’s a weekly overview:

As you can see, the portfolio is currently in a drawdown so I’m looking forward to seeing if we get a rebound by the end of the quarter. I’m expecting big things on September 17th. According to my alert system, this is the highest confidence trade date for the portfolio this year. I’ll talk more about why that is in a moment.

Q3 Performance

Looking at the results of the Forward Test by day of week, the best days were Monday (+$69,545) and Tuesday (+$54,775). The worst day of the week was Friday (–$63,175). Wednesday had the highest win rate (64.6%), but still lost money (–$12,963). Key takeaway: This regime is deadly for strategies on Friday. Feast on Monday and Tuesday, however.

The belles of the ball:

• Strategy 42: $50,875 (just 45 trades)

• Strategy 3: $18,615

• Strategy 77: $53,988 from 23 trades

Scroll to the bottom of this post for the full forward test.

Read more

For links to all strategies click here

We're not just developing strategies—we're on a quest for the holy grail of automated trading. Questions? Check the FAQs first, then feel free to reach out directly: AutomatedTradingStrategies@protonmail.com.

The evolution of the Forward Test:

• 2023: After years of failed live tests based solely on backtests and walk-forward analysis, I finally started forward testing. It was a game changer.

• 2024: Methodology upgrade—began tracking individual variations instead of strategy groups. Allowed for greater granularity and deeper insights.

• 2025: Alpha hunting through ML pattern recognition, optimal portfolio allocation, and advanced risk management techniques. Using an Incubator portfolio to scout strategies for inclusion in the next Forward Test. The Incubator is currently running over 200 strategies. I’ll share an update on that portfolio in a few days.

What is the Q3 Portfolio Composition?

• The Q3 Forward Test is the “production/candidate” portfolio. Every strategy here has already cleared the Incubator hurdles. The forward test uses live data on simulated accounts. The goal is to run a capital‑efficient, broadly diversified, but tightly risk‑capped portfolio that can be migrated to a live account.

• The Q3 Forward Test is a 20‑strategy, multi‑asset portfolio. It netted $63K on 884 trades executed between late June and July 18, 2025 with a 2.0 Profit Factor—driven mainly by gold (GC). This portfolio corresponds with the September 2025 equity contract.

Instruments Traded:

• NQ (Nasdaq-100 E-mini) - Dominates with momentum and mean reversion strategies

• ES (S&P 500 E-mini) - Regular positions

• GC (Gold) - Precious metals are on fire! Longer hold times, often overnight positions. 74 % of portfolio profit is coming from three GC strategies that together generate only 10 % of the trades—great edge. Lots of action in metals this quarter.

• ZN/ZB (Treasury Notes/Bonds) - Fixed income futures. One strategy exclusively trades bonds with 5-contract positions. May be crowding margin that could fund better candidates.

• MNQ (Micro Nasdaq) - Smaller position sizing

Trading Characteristics:

• High Frequency: Some strategies trade multiple trades per day

• Mixed Timeframes: From 5-minute to 90-minute strategies

• Both Long and Short: Strategies take positions in both directions

Position Sizing

• Most trades use 1-3 contracts

• Some strategies scale up to 5-8 contracts

• Position sizing is correlated with strategy confidence and/or market conditions. I’ll be introducing a few descriptive metrics to help sort this out a bit.

Exit Strategies

• Profit targets: Most common exit method

• Trail stops: Used for trend-following

• Exit on session close: Risk management for overnight exposure

• Time-based exits: Some positions closed after specific durations

Performance

Big drawdowns at the start of the quarter (down nearly $20,000), but the system recovered to a profit of $62,890 by July 18. An overview of portfolio performance by strategy, trade count, profit, profit factor, win rate % and dollars/trade is at the bottom of this post.

In the Pipeline…

Incubator Strategies: There are many promising strategies in the Incubator. Some have been published, some haven’t. Look for a post in a few days that takes a closer look at these strategies. I’ll also introduce some new descriptive metrics in this post. See below for a quick look at the top 10 strategies in the Incubator by net income.

The Incubator includes AI strategies, as well as a strategy created by the latest model from Grok—Grok 4. Grok is one of four models chosen by the the Pentagon to help the U.S. military expand its use of advanced artificial intelligence capabilities. In the same way that we’re using AI to create agentic workflows for trading, the Pentagon is using it to develop agentic AI workflows for national security. The other models were from Google (Gemini), Anthropic (Claude) and of course OpenAI (ChatGPT).

Alerts: RL model alerts will start in September. Every subscriber can receive an alert for the two strategies selected for your Forward Test. You can read more about that process here.

August: I will be away from my desk for the month of August. All subscribers as of July 31 will receive one month free added to your subscription. This will be automatic. No action is required on your part, however, if you don’t see the adjustment, please send me an email. I have a few posts in the queue for August, but will not respond to email until I return in September.

If you have any questions, please send me an email.

Contact: Celan @ AutomatedTradingStrategies@protonmail.com

For links to all strategies click here

Q3 Performance and Top 10 Incubator Strategies

Read more

For links to all strategies click here

We're not just developing strategies—we're on a quest for the holy grail of automated trading. Questions? Check the FAQs first, then feel free to reach out directly: AutomatedTradingStrategies@protonmail.com.

The evolution of the Forward Test:

• 2023: After years of failed live tests based solely on backtests and walk-forward analysis, I finally started forward testing. It was a game changer.

• 2024: Methodology upgrade—began tracking individual variations instead of strategy groups. Allowed for greater granularity and deeper insights.

• 2025: Alpha hunting through ML pattern recognition, optimal portfolio allocation, and advanced risk management techniques. I am also creating a model that uses RL to train data at the strategy/trade level for strategy selection. The reward is a good prediction. The reward is increased for a higher profit factor. Strategy selection is on a daily basis and the model learns from each new day.

What is the Q2 Portfolio Composition?

• Scale: 11 distinct strategies with 37 variations

• Timeline: Launched March 23, 2025, 6 PM EST. Will run through mid-June.

• Legacy performers: Three strategies carried over from Q1 (4c, 44b, 77c).

• New blood: Eight strategies advanced based on stellar Q2 Incubator performance.

• Instrument Diversity:

  • NQ (Nasdaq futures): 22 strategies

  • ES (S&P 500 futures): 7 strategies

  • MNQ (Micro Nasdaq futures): 2 strategies

  • GC (Gold futures): 2 strategies

  • Specialized instruments: ZB (Treasury bonds), RTY (Russell 2000), CL (Crude oil), 6B (British Pound), YM (Dow Jones), and HG (Copper)

Timeframe Range: Multiple timeframes from 3 to 90 minutes

So how is the portfolio doing over the last 11 weeks? We’ll look at:

• Top 5 performing Strategies

• Top 5 performing hours by profit factor for the portfolio (these are your best hours to enable strategies)

• Worst days and hours to enable strategies in the portfolio

• Sequential patterns and structural edges found with ML techniques along with three strategy ideas to take advantage of these patterns.

If you’re just looking for a chart with the performance results (profit factor, trade count, win rate, and net profit) of all strategies in the portfolio, please scroll to the bottom.

Read more

We're not just developing strategies—we're on a quest for the holy grail of automated trading. Questions? Check the FAQs first, then feel free to reach out directly: AutomatedTradingStrategies@protonmail.com.

*Not all strategies are included with the subscription. Click the button below for a list of all strategies included with the subscription.

For links to all strategies click here

The evolution of the Forward Test:

• 2023: After years of failed live tests based solely on backtests and walk-forward analysis, I finally started forward testing. It was a game changer. Backtests can give you a general indicator of edge, but they often fail, especially for intraday strategies. Click here for a list of reasons why backtests often differ from actuals—and this just talks about the engine, not the data. Your best bet is to run as many strategies in a forward test as you can. Use the data stream you will use when trading live. Automate the entire process and then identify/analyze the winners and losers. That’s what I’m doing here.

• 2024: Methodology upgrade—began tracking individual variations instead of strategy groups. Allowed for greater granularity and deeper insights.

• 2025: Alpha hunting through ML pattern recognition, optimal portfolio allocation, and advanced risk management. Strategy 94 was the first application of this, and I’ll provide an update on how Strategy 94 is performing in a moment.

What is the Q2 Portfolio Composition?

• Scale: 11 distinct strategies with 37 variations

• Timeline: Launched March 23, 2025, 6 PM EST. Will run through mid-June.

• Legacy performers: Three strategies carried over from Q1 (4c, 44b, 77c).

• New blood: Eight strategies advanced based on stellar Q2 Incubator performance.

• Instrument Diversity:

  • NQ (Nasdaq futures): 22 strategies

  • ES (S&P 500 futures): 7 strategies

  • MNQ (Micro Nasdaq futures): 2 strategies

  • GC (Gold futures): 2 strategies

  • Specialized instruments: ZB (Treasury bonds), RTY (Russell 2000), CL (Crude oil), 6B (British Pound), YM (Dow Jones), and HG (Copper)

Timeframe Range: Multiple timeframes from 3 to 90 minutes

So how is the portfolio doing?

Key Insights:

• Portfolio is green: full log shows +$30.4K net on 1,199 contracts, but the win is fragile.

• Eighty-five percent of profit comes from 9 small-lot strategies (each >$1K, none ever >3 contracts). They account for just 35% of total risk.

• Time risk: the worst draw (-$52K) happened trading ES around 10:30 a.m. ET; best burst (+$236K) was NQ shorts held past the 4 p.m. cash close.

• Edge window: heat-map confirms Mon–Wed 4:00-09:00 ET and Thur afternoon fade are positive; Fridays show mild bleed.

• Hedging isn’t a buzzword. The ZB play proved its worth exactly when ES blew up.

Top five strategies this quarter:

1. SimAccount77 — “Deep-Swing”

Stat line: 12 contracts, +$86,060, Profit Factor ∞ (zero losers), Max MAE $6,905

Think of this one as the swing-trader who shows up twice a week, nails the direction, and leaves before the crowd realizes what happened. The logic is built around a 74-minute bar series—an oddball timeframe that filters out micro-noise, but still catches asymmetrical overnight moves. Twelve entries, twelve exits, all green—hence the “infinity” profit factor. Note: This strategy is currently unavailable with the subscription. Please contact me directly if you would like to purchase.

2. SimAccount93 — “The Siege Engine”

Stat line: 44 contracts, +$236,965, Profit Factor 4.5, Max MAE $47,970

This is the high-conviction discretionary overlay that drops two- and three-lot mini-swings when Nasdaq sentiment turns. The monster win occurred on April 4 from a cascading short that captured a 900-point intraday flush. The real talent here is discipline on re-entry: after every scale-out the system demands a fresh 30-minute lower-high before adding size. That rule alone cut what would have been a $140K drawdown in half. I had to disable this strategy due to high volatility, but will be adding its tweaked cousin to the Incubator in Q3.

3. SimAccount43 — “Mean-Reverting Baby”

Stat line: 2 contracts, +$4,620, Profit Factor 13.2, Max MAE $745

A boutique filter—this is the “catch-the-knife safely” play: microscopic trade count, laser-tight stops, and a one-contract scale-out. In back-tests it fires six times a month; in real trading I’ve seen exactly two fills. This strategy is available for download here.

4. SimStrategy49 — “Risk-Off Hedge”

Stat line: 42 contracts, +$6,281, Profit Factor 3.2, Max MAE $3,938

Whenever ES volatility spikes, this bond-futures strategy quietly goes long three or four ZB contracts. It exits in thirds. The kicker: during the ES 52-min disaster on May 2, ZB booked +$1,406, shaving 7% off the day’s portfolio draw. This strategy is available for download here.

5. SimAccount94 — “Nano-Scalper Deluxe”

Stat line: 83 contracts, +$8,305, Profit Factor 2.1, Max MAE $3,060

This is your high-frequency crowd-pleaser —ninety-second average hold, 100% win-rate. It lives for the first wave of US cash hours (09:30-11:30 ET). The edge isn’t massive per trade (about $100 gross) but at 40–50 micro-fills a week the wins snowball without scaring the risk committee. The problem with crowd-pleasers is that they tend to be undercover alchoholics—every now and then they crash. When this happens, the impact on the portfolio is brutal. While this variation performed well, other variations did not. The good news is that I’ve been testing several variations of this strategy and one variation is the clear winner. I’ll share that analysis with you in a moment. Strategy 94 is available for download at the bottom of this post.

For links to all strategies click here

With that, I want to share with you the full forward test. Again, this is not a backtest. This is a forward test using simulated accounts on live data.

The first column is the Strategy, followed by the number of trades, the approximate strategy start date, the date of the last trade, net profit and finally maximum MAE (intra-trade drawdown). Note, the first and last strategies have been disabled since April 4 due to extreme volatility as evidenced by the Max MAEs for both strategies:

Read more

We’re on the hunt for the holy grail of automated trading strategy. If you have any questions, start with the FAQs and if you still have questions, feel free to reach out to me (Celan) directly at AutomatedTradingStrategies@protonmail.com.

For links to all strategies click here

“Markets are rarely random, they are just good at hiding their patterns.”
— Jim Simons (2024 Bloomberg New Economy Forum)

As we tap into the realm of machine learning to discover hidden market patterns, we are entering a new part of the forest. It may be uncomfortable at times, but the rewards are great.

Let’s go!

From Finding the Watering Holes to Building an Oasis

Strategy 96 is a continuation from Strategy 95. It is one of my favorite strategies because we use ML to find non-linear patterns that we always knew were there.

The goal of Strategy 95 was to see if we could improve upon a simple MACD cross strategy using market regimes identified via clustering. We found some really nice watering holes. Now it’s time to see how long they persist.

Recap: What Strategy 95 Taught Us

• Clustering is great, but not a crystal-ball, it’s a map showing which local terrains (regimes) have historically been lush.

• We can create clusters based on any metric. Our focus is on high-win-rate clusters (Steady Oasis, Mr Precision, Dynamic Optimist). These clusters repeat across bar sizes and instruments.

• Simply flipping, throttling, or sitting out the MACD in unfavorable clusters already improves the profit factor vs. vanilla MACD.

We Know Where—Now Let’s Ask How Long: Why Persistence Matters (and how to measure it)

“Edge isn’t just where you trade; it’s when you trade the most.”
— Linda Raschke (Traders’ Chat, May 2025)

Strategy 96 is our first well-digging exercise.

1. Identify a lush spot (cluster).

2. Measure how long it persists.

3. Size the trade accordingly.

Clusters say “this bar is green.” They don’t say “green will probably last another seven bars.” In 96, this is the question we want to answer. We are looking for the probability that green will ‘persist’.

It’s important to distinguish between accuracy (win rate) and persistence. Just because a regime is persistent does not mean that it has a high win rate or vice versa. A regime may produce a high win rate for only one bar. Likewise, a regime can be persistent, lasting 7 or 8 bars, but have a low win rate. Best case scenario is a highly profitable regime with a high degree of persistence. That’s what we’re looking for in Strategy 96.

Strategy 96 uses six market regimes. The six regimes come from a k-means clustering run (k = 6) on the standardized 30-minute-bar feature matrix. They are as follows:

I then measured the performance of each regime by profit factor. An extremely good argument can be made for using win rate given the nature of the command (MACD cross up) and I plan on adding a version of that strategy to the Incubator. Meanwhile, here’s the impact of toggling each regime by profit-factor.

Regimes 3 and 4 are the heavy lifters—turning them on boosts profit factor an order of magnitude compared with runs where they’re off. Regimes 0-2 act like noise filters—keeping them off correlates with much higher PF.

So we’re going to enable Regimes 3 and 4 because they have the highest profit factor.

Now let’s measure the persistence of Regimes 3 and 4.

To measure persistence, I quantified how long Regimes 3 & 4 normally last (e.g., R3 holds 2 bars ≈ 100 %, 3 bars ≈ 78 %), and then converted those odds into rules. Here’s a transition matrix for Regimes 3 and 4:

R3 is sticky (85% hold rate); when it flips, most often to R2 (8%). R4 is less stable—60% chance to stay, 25% chance it degrades into choppy R2.

How can we use this information to size?

So the dynamic sizing is as follows:

• BaseQty input sets your normal size.

• Regime 3 ups size by +1 after two bars and +2 after three.

• Regime 4 ups size by +1 after two bars.

• All other regimes revert to BaseQty.

How does Strategy 96 use this data?

Read more

We're not just developing strategies—we're on a quest for the holy grail of automated trading. Questions? Check the FAQs first, then feel free to reach out directly: AutomatedTradingStrategies@protonmail.com.

*Not all strategies are included with the subscription. Click the button below for a list of all strategies included with the subscription.

For links to all strategies click here

Despite significant market volatility, the portfolio of 11 core strategies (31 variations across multiple instruments like NQ, ES, GC) executed 491 trades for a net profit of $49,800.25.

Key Performance Highlights:

• Top Net Contributors: The "77" momentum strategy variations collectively added $97,296.50.

• Consistent Performer: The "94" high-frequency strategy variations ("Precision Sniper" leading) contributed $18,605 with a high win rate (96.2%), proving valuable for consistency.

• Diversification: Strategies on exotic instruments (ZB, RTY, CL, GC, etc.) added $4,808.75 and provided balance during turbulence.

In this overview, I’m also going to look at several rehabilitation strategies including exit timing optimization and equity curve analysis to improve performance. The goal is continuous improvement.

Let’s get started.

The Champions: Strategy Leaderboard

This Q2 Forward Test update covers the first three weeks of trading data across multiple accounts, strategies, and instruments. It uses live data on simulated accounts. By examining the performance metrics, timing, and behavior of each strategy, we can extract valuable insights to optimize our approach going forward.

These are the current Mudders. I’ve given them temporary nicknames that capture distinctive characteristics within each strategy. Scroll to the bottom for a key that matches the nickname to the account name used in the Forward Test.

All strategy variations with an orange marker are not included with the subscription. For links to all strategies included with the subscription click on the button below.

For links to all strategies click here

Let's dissect our champion strategies—not just what they do, but how they do it. These aren't theoretical constructs. They aren’t backtests. These are forward tests based on live data run on simulated accounts.

Q2 2025 Portfolio Composition

• Scale: 11 distinct strategies with 31 variations

• Timeline: Launched March 23, 2025, 6 PM EST running through mid-June

• Legacy performers: Three strategies carried over from Q1 (4c, 44b, 77c)

• New blood: Seven strategies advanced based on stellar Q2 Incubator performance.

• Instrument Diversity:

  • NQ (Nasdaq futures): 22 strategies

  • ES (S&P 500 futures): 7 strategies

  • MNQ (Micro Nasdaq futures): 2 strategies

  • GC (Gold futures): 2 strategies

  • Specialized instruments: ZB (Treasury bonds), RTY (Russell 2000), CL (Crude oil), 6B (British Pound), YM (Dow Jones), and HG (Copper)

Timeframe Range: Multiple timeframes from 3 to 90 minutes

Scroll to the bottom of this page for a link to the Q2 2025 Portfolio and full Q2 results. For help with creating your own forward test, click here and feel free to post questions in the comments below.

Strategy Profiles

The Maestro and the Orchestra

The portfolio's undisputed maestro, The Overnight Oracle, conducted an impressive solo performance, generating a staggering $236,965 in profits across 44 trades. That said, the strategy is highly volatile and has huge drawdowns, so I’ll be pausing until I can rework.

Supporting the maestro was the talented "77" quartet: The Momentum Whisperer, The Trend Surfer, The Wave Rider, and The Micro Maestro. These four strategies share the same core strategy, but are all variations on a momentum theme playing at different frequencies (timeframes), harmonized beautifully to add $97,296.50 to the bottom line in just the first three weeks of Q2. The Momentum Whisperer's 74-minute timeframe proved most lucrative, capturing the sweet spot between noise and predictability.

For links to all strategies click here

The Rhythm Section

While the headliners commanded attention, the rhythm section—the "94" strategy family—maintained a steady beat throughout. These six variations of high-frequency, precision-targeted approaches executed 184 trades with an astonishing 96.2% win rate. The Precision Sniper led this section with its 18-minute timeframe precision, demonstrating that consistency and mathematical edge can be just as valuable as occasional brilliance.

The beauty of this rhythm section lies not in individual trades, but in the cumulative effect of their persistence—like raindrops eventually filling a bucket, these small but consistent winners added $18,605 to the portfolio. You can download Strategy 94v2 below, which includes four embedded scenarios to auto-fill parameters.

This strategy is based on Strategy 44b. After deep analysis of time-of-day patterns and sequential trade behaviors, Strategy 94 emerged as the enhanced version of its predecessor—skipping the Incubator phase to debut directly in the Q2 Forward Test. Note: only 6 of the 10 variations have been profitable so far, so I’ll be disabling the four variations that were unprofitable.

The Exotic Instruments

Adding color and diversity to this symphony were the specialty market instruments—The Safe Harbor in treasury futures, The Swift Tactician in Russell futures, The Bullion Baron and The Oil Tycoon. Though contributing a modest $4,808.75, these strategies provided crucial diversification during market turbulence, often moving counter to equity-based strategies. All of these strategies are available with the subscription.

The Finale

In total, this trading symphony performed 491 trades for a net profit of $49,800.25 during one of the most volatile periods in recent market history. The performance revealed a portfolio with a mixture of momentum-based approaches generating 88% of profits, precision scalping providing consistency, and specialty markets offering balance.

Like any great composition, the whole proved greater than the sum of its parts. The strategies complemented each other across different market conditions, timeframes, and instruments, creating a robust approach that could navigate even the most volatile conditions.

Looking ahead to Week 4 & 5, the focus turns to optimizing this symphony—increasing allocation to star performers, rehabilitating struggling artists, and fine-tuning the interplay between strategy families. The difference between ruin and reward is often a single adjustment. An inch here, an inch there can turn a loss into a win.

One rehabilitation measure that I believe will bear real fruit is optimizing ETD.

Optimizing Exit Techniques: The ETD Opportunity

One of the most significant findings from my analysis involves the substantial room for improvement in exit timing, as evidenced by the ETD metrics.

Read more

We're not just developing strategies—we're on a quest for the holy grail of automated trading. Questions? Check the FAQs first, then feel free to reach out directly: AutomatedTradingStrategies@protonmail.com.

*Not all strategies are included with the subscription. Click the button below for complete list of all strategies included with the subscription.

For links to all strategies click here

The recent price action in the Nasdaq-100 futures (NQ) deserves close attention. After nearly two years of steady climbing within a well-defined upward channel, we're seeing a significant technical breakdown that could signal a shift in market dynamics.

Looking at the daily chart spanning from early 2023 through March 2025, the NQ had established a remarkably consistent uptrend, bounded by parallel trendlines that contained price action through multiple rallies and corrections. This channel acted as guardrails for traders navigating the index, but those guardrails were violated in early March.

It is worth noting that this correction began as a typical retracement from the upper boundary and quickly accelerated into a more substantial decline, eventually breaching the lower support trendline that held firm for nearly two years.

Most telling is what happened next: rather than immediately bouncing higher, price retraced upward just enough to test the underside of the broken trendline before selling resumed. This "kiss goodbye" pattern, where former support transforms into resistance, is a classic bear signal that often confirms a meaningful trend change rather than a temporary correction.

The technical indicators support this bearish narrative, with momentum oscillators showing pronounced weakness despite the brief retracement attempt. This suggests that selling pressure remains dominant even after the initial breakdown.

What does this mean for us traders?

The violation of such a long-standing uptrend channel indicates we may be entering a new market phase that requires adjusted strategies. The failed retest of the trendline provides a clear reference point for risk management, while also suggesting that the path of least resistance may now be downward rather than upward. I’ve added a green trendline to the chart below to show the downtrend.

This technical breakdown doesn't necessarily signal doom, but rather a potential regime change that warrants closer attention to position sizing and sector exposure. After an extended uptrend of this sort, a period of consolidation or deeper correction would be historically consistent with healthy market behavior. Whether this breakdown marks the beginning of a more extended downtrend or merely sets the stage for a more complex bottoming process remains to be seen. What's clear is that the easy, channel-bound uptrend that defined the NQ for nearly two years now appears to be undergoing a significant test.

Why is this happening: tariffs, changes in political regimes, DOGE? I get all my news from comedians so what do I know, but one thing is certain—we’ve got major headwinds and tailwinds in focus.

Market Headwinds

1. Technical breakdown: As we just discussed with the NQ chart, the breach of a long-term uptrend channel suggests potential technical weakness ahead.

2. Valuation concerns: Many tech stocks and market indices are trading at historically high multiples, creating vulnerability to earnings disappointments.

3. Interest rate environment: While rates will eventually come down (fiscal policy leads monetary policy for once), current elevated rates create pressure on growth stocks and increase borrowing costs for companies and consumers.

4. Geopolitical tensions: Ongoing conflicts and trade tensions create uncertainty that markets typically don't appreciate.

5. Recession fears: Despite resilience so far, concerns about economic slowdown persist in some sectors.

Market Tailwinds

1. AI and technological innovation: Continued advancements in AI, automation, and other technologies drive productivity and new business opportunities to new heights. I don’t think the world is ready for the unprecedented productivity gains we’re about to see over the next 2 years.

2. Corporate profitability: Many companies continue to report solid earnings despite challenging conditions.

3. Consumer strength: Employment remains robust and consumer spending has shown resilience.

4. Potential monetary easing: Expectations for future rate cuts could provide support for risk assets.

5. Institutional positioning: Dry powder is still at record levels. Many funds still have cash to deploy, potentially providing support on significant pullbacks.

The interplay between these opposing forces will likely determine market direction in the near to medium term.

Where are we on the Q2 Forward Test after last week?

I know it’s only been one week, but I wanted to provide a quick update on the Q2 Forward Test (scroll to the bottom for a link to the portfolio). It is composed of 11 distinct automated trading strategies with 37 variations. Only 14 of those 37 variations were triggered last week.

For links to all strategies click here

I don’t normally like to make changes once the Forward Test has started, but I’ve created an agentic portfolio manager that provides recommendations/suggestions on changes to both composition and allocation so I’ll be implementing some of those changes this quarter.

For those of you just joining us, here’s a quick overview of the Q2 forward test composition which uses live data on simulated accounts:

• Timeline: March 23, 2025, 6 PM EST through mid-June

• Legacy performers: Three strategies carried over from Q1 (4c, 44b, 77c)

• New blood: Eight strategies advanced based on stellar Q2 Incubator performance

• Instrument Diversity:

  • NQ (Nasdaq futures): 22 strategies

  • ES (S&P 500 futures): 7 strategies

  • MNQ (Micro Nasdaq futures): 2 strategies

  • GC (Gold futures): 2 strategies

  • Specialized instruments: ZB (Treasury bonds), RTY (Russell 2000), CL (Crude oil), 6B (British Pound), YM (Dow Jones), and HG (Copper)

• Timeframe Range: Multiple timeframes from 3 to 90 minutes

• Recommended account size: $100K

Total net profit for the first week of the quarter was $28,802 on 115 trades.

Several strategies tanked and I’m going to pause them for now, but several did quite well, including one that made $81K and another that continues to have a 100% win rate trading CL. Here’s a chart with strategy performance (trade count, total profit, win rate, avg profit, avg loss) for the first week in the quarter:

Read more

We're not just developing strategies—we're on a quest for the holy grail of automated trading. Questions? Check the FAQs first, then feel free to reach out directly: AutomatedTradingStrategies@protonmail.com.

Quick reminder: All seven strategies from the Q1 Forward Test have been ‘unpublished’ and are no longer available with the subscription. Only three strategies survived Q1 to join the Q2 cohort. After June's Q2 completion, additional strategies will be removed. All subscribers will receive advance notice before any removals.

Need links to all subscription strategies? Click below for complete list.

For links to all strategies click here

Call to Arms: Your Forward Test Results Wanted

Friends, countrymen, traders, lend me your forward test results!

I know many of you want to help. Over the years I’ve come to realize that the most valuable contribution you can make is sharing forward test results. And if you have strategy improvement suggestions, I’d love to hear them, but please validate them with forward test data first. Real-world performance speaks louder than theoretical improvements. I look forward to hearing from you!

Q2 2025: Unleashing the Beast

We’ve learned a lot over the last four years, and even more over the last four quarters. I want to thank everyone for all your feedback and support. In particular, I want to thank Jeff, Xa, Kevin, Z and Cora. Your support is greatly appreciated.

So where are we today…

Today I’m publishing the Q2 2025 Forward Test. This is the culmination of years of hard work and is by far the most ambitious portfolio I’ve ever ran.

In the last Mudder Report, we separated the Q1 winners from losers and introduced Strategy 94—featuring 10 variations for Q2. While not traditionally forward tested, Strategy 94 evolved from one of our star performers: Strategy 44b.

Strategy 44b's performance record speaks volumes:

• Crushed expectations in Q3 and Q4 of 2024

• Continued domination through Q1 2025

• Extraordinary stats: only 2 losing trades out of 79 total trades

• Both losses occurred at midnight on 2/21 (pattern identified and addressed)

• average trade duration is 7 min allowing for increased capital allocation

After deep analysis of time-of-day patterns and sequential trade behaviors, Strategy 94 emerged as the enhanced version of its predecessor—skipping the Incubator phase to debut directly in the Q2 Forward Test. So while Strategy 44b is no longer available through the subscription, you can download its successor here.

Q2 2025 Forward Test Launch

The evolution of the Forward Test:

• 2023: Initial serious commitment to live forward testing

• 2024: Methodology upgrade—tracking individual variations instead of strategy groups

• 2025: Next-level alpha hunting through ML pattern recognition, optimal portfolio allocation, and advanced risk management. Strategy 94 is the first application of this.

Q2 Portfolio Composition

• Scale: 11 distinct strategies with 37 variations

• Timeline: Launching today (March 23, 2025, 6 PM EST) through mid-June

• Legacy performers: Three strategies carried over from Q1 (4c, 44b, 77c)

• New blood: Eight strategies advanced based on stellar Q2 Incubator performance

• Instrument Diversity:

  • NQ (Nasdaq futures): 22 strategies

  • ES (S&P 500 futures): 7 strategies

  • MNQ (Micro Nasdaq futures): 2 strategies

  • GC (Gold futures): 2 strategies

  • Specialized instruments: ZB (Treasury bonds), RTY (Russell 2000), CL (Crude oil), 6B (British Pound), YM (Dow Jones), and HG (Copper)

Timeframe Range: Multiple timeframes from 3 to 90 minutes

Recommended account size: $100K

Incubator Standouts Join the Field

Three of the strategies in the portfolio are from the Q1 Forward Test (4c, 44b and 77c). All other strategies in the Q2 Forward Test were advanced based on their performance in the Q2 Incubator.

The Q2 Incubator showcases different variations of the first 50 strategies. In Q3, I’ll add strategies 51 through 94 and in Q4, I’ll add all strategies with embedded contracts. Here’s a tentative timeline:

The hope is to obtain as much actual data as possible for ML pattern recognition analysis by year end.

I’ve only been running the Q2 Incubator for one month, but there have already been some very promising results.

Read more

Housekeeping:

We’re on the hunt for the holy grail of automated trading strategy. If you have any questions, start with the FAQs and if you still have questions, feel free to reach out to me (Celan) directly at AutomatedTradingStrategies@protonmail.com.

As we continue on this journey, the strategies are becoming more complex in nature. I’ll be making some changes going forward to help cut back on the configurable parameter set. Meanwhile, please feel free to ask questions and/or provide feedback about ways to make the process easier.

***As a quick reminder, all 7 strategies in the Q1 Forward Test have been unpublished. I’ll be removing more strategies after completion of the Q2 Forward Test. All subscribers will be notified prior to removal. For links to all strategies included with the subscription, click on the button below.

For links to all strategies click here

The Hunt for Alpha: Q1 2025 Forward Test Results

We just finished week 10 of the Forward Test, which means we’re at the end of Q1. What a wild quarter!

I’m going to show you which strategies failed the test and which passed with flying colors. I’m also going to share Strategy 94 with you. It produced the following backtest:

This strategy is based on an amazing pattern I found in Strategy 44b’s trade data using machine learning algorithms. Strategy 94 takes the pattern I found and uses it to trade greater size when the pattern is confirmed. I can’t wait to share this strategy and its feature with you.

Before I do, let’s review some key insights found in the Q1 Forward Test that can help with the Q2 Forward Test.

2025 Q1 Forward Test Results

The Q1 Forward Test:

• Is built from 7 core strategies and composed of 19 battle-tested strategy variations

• Uses the following instruments: ES, NQ, HG, MNQ, ZB, YM, and RTY

Scroll to the bottom of this post for a look at the final Q1 numbers for each strategy variation, including net profit, profit factor, and trade count. You can also compare Q1 against Q4 of 2024.

The Evolution of The Forward Test

• 2023: Started getting serious with live forward testing

• 2024: Leveled up by tracking each variation (no more lumping strategies together).

• 2025: The next step is to understand how we can gain more alpha through ML pattern recognition, portfolio allocation and risk management. The forward test documents this effort.

And I want to be clear, the following results are not from a backtest. These are the results of the final results for the 2025 Q1 Forward Test, which uses simulated accounts on live data.

Let’s get started…

Q1 2025 Forward Test Results

This is where we left off with the last Mudder Report after 5 weeks of data:

Read more

For links to all strategies click here

I know we’ve had an influx of new subscribers over the last month. So as a quick reminder, we’re on the hunt for the holy grail of automated trading strategy. If you have any questions, start with the FAQs and if you still have questions, feel free to reach out to me (Celan) directly at AutomatedTradingStrategies@protonmail.com.

The Nobel Prize in Chemistry was divided in 2024: one half was awarded to David Baker "for computational protein design"; the other half jointly to Demis Hassabis and John Jumper "for protein structure prediction".

Demis Hassabis is also Google Deepmind’s CEO, and he gave a fascinating lecture in acceptance of the award. The title of the lecture: "Accelerating scientific discovery with AI".

This is the slide that caught my eye.

This slide is showing three criteria that make a problem well-suited for AI:

• Massive combinatorial search space

• Clear objective function (metric) to optimize

• Either lots of data and/or an accurate and efficient simulator

While hardly as noble as computational protein design or protein structure prediction, the requirements map directly onto automated/algorithmic trading: lots of possibilities to explore, a clear objective (returns), and abundant data.

Financial markets present a huge range of possible trades. AI systems can sift through countless permutations—across various instruments, time frames, and economic conditions. Likewise, traders typically have a very clear objective—maximize returns, minimize losses. This provides an explicit signal that an AI can optimize against. And we have so much data it’s hard to find a file format that can handle it all.

The simulator part is a work-in-process, however. I’m currently testing the use of machine learning models to develop strategies based on simulated market data and then verifying the answer with actual data, which triggers an adjustment to the simulation. The idea was inspired by Grok 3, which used simulated data to train on, and actual data to verify against.

So the good news is that our hunt is a natural application for AI methods. The challenge is in the application.

Applying ML To The Strategy Creation, Selection & Improvement Process

At the beginning of 2024 I started my first Live Test. I hope to conduct another, but this year I wanted to devote more time to studying the application of ML to trading. I believe it has the ability to take our hunt to new heights.

Read more

For links to all strategies click here

As a quick reminder, we’re on the hunt for the holy grail of automated trading strategy. If you have any questions, start with the FAQs and if you still have questions, feel free to reach out to me (Celan) directly at AutomatedTradingStrategies@protonmail.com.

The Hunt for Alpha: Q1 2025 Forward Test Results

Alright traders, we just finished week 5 in a 10 week quarter. Here’s a performance chart of cumulative net income over the test period.

We're not playing with backtests and hypotheticals. This is live NT8 data, running in real-time through simulated accounts. Why does that matter? Forward testing separates the dreamers from the doers. Does it guarantee future performance? Hell no—nothing in trading does—but it puts the odds in your corner.

Here's what you need to know about the Q1 2025 forward test:

The Arsenal

• 19 battle-tested strategy variations

• Built from 7 core money-makers

• All running on 30+ minute charts

• Instruments of choice: ES, NQ, HG, MNQ, ZB, YM, and RTY

The Evolution of The Forward Test

• 2023: Started getting serious with live forward testing

• 2024: Leveled up by tracking each variation (no more lumping strategies together). That led us to a few Champions that now comprise the Q1 2025 Forward Test Portfolio.

• 2025: The next step is to understand how we can gain more alpha through time based strategy selection, portfolio allocation and risk management. This Mudder Report is a first step in that effort.

So today we're answering the questions every trader asks, but few can answer with data:

• "Got one day to trade? Here's when to show up."

• "Only got two hours? Here's when to deploy."

• "Can only run three strategies? Here's your weapons of choice."

We're about to dive into:

• Raw performance numbers

• Top guns by net profit

• Strategy stacking opportunities

• Negative correlations (perfect pairs trades)

• Days that pay & time windows that print (when to push the boat out)

Ready to see where the real edge is hiding? Let's dive in...

Overall Performance

• Total Trades: This is an active, intraday system. There are 113 trades in the period. That’s an average of roughly 3–4 trades per day.

• Net Profit & Average Trade: The final net profit at 113 trades is $84,061. That works out to roughly $84,061 ÷ 113 ≈ $743 per trade on average.

  

• Win/Loss Breakdown: 67% win rate (76/113 ≈ 67%) and a 33% loss rate.

• Instrument Mix: The bulk of the trading is on NQ (NASDAQ E‐mini) and its cousin MNQ (micro NQ). There are also trades in ES (S&P 500 E–mini), YM (the Dow futures), RTY (Russell 2000), HG (Copper) and ZB (10yr)

  Strategy Performance by Net Profit:

Read more

Four Years Later: From Holy Grail to Reality

Four years ago, on January 14, 2021, I embarked on what seemed like an impossible quest: creating the holy grail of automated trading strategies. Our benchmark criteria were ambitious, to say the least:

• Profit factor > 3 (At the time, this seemed as likely as finding a unicorn in your backyard)

• Annual drawdown < 3%

• Annual return on max drawdown > 500%

• Maximum daily net loss limited to $1,000

• Average daily profit > $1,000

• Annual trade count: 253-5,000 trades

While we didn't find a single strategy that could turn water into wine, we did something potentially more valuable: we built a portfolio of strategies that not only met but exceeded these criteria. And instead of just using the backtest, we validated with a forward test. Click on the link at the bottom of this post for a look at how the strategies in the Q1 2025 Forward Test portfolio performed over the last two quarters delivering:

• $84K profit, 3.57 profit factor in Q3 2024; and,

• $289K profit on 577 trades, 11.14 profit factor in Q4 2024.

Q1 2025 Forward Test Portfolio Composition

A key lesson from 2024: tracking individual strategy variations rather than aggregate performance was a game-changer. The current portfolio comprises 7 strategies with 19 variations, all trading on 30-minute charts or higher across a carefully selected basket of futures contracts.

Instrument selection reflects a strategic approach to market exposure:

Equity Index Futures

• ES (S&P 500 E-mini): Our core large-cap market benchmark

• NQ (Nasdaq-100 E-mini): Technology-weighted exposure

• YM (Mini Dow): Industrial-focused blue-chip exposure

• RTY (Russell 2000 E-mini): Small-cap market segment

• MNQ (Micro Nasdaq-100): Enables finer position sizing and risk management

Diversification Instruments

• HG (Copper): Often called "Dr. Copper" for its economic forecasting properties

• ZB (30-Year Treasury Bond): Provides fixed-income exposure and potential hedge against equity market stress

All timeframes are 30-minutes or higher—these intervals offer an optimal balance between noise reduction and opportunity capture. Shorter timeframes (5 min, 15 min) often generate excessive noise and trading costs, and the selected timeframes provide cleaner signals while still capturing significant market moves.

19 variations across 7 strategies provide robust diversification benefits:

• Parameter Diversity: Each variation uses slightly different entry/exit parameters, allowing us to capture similar edges under different market conditions

• Risk Management: Variations help spread risk across multiple approaches rather than relying on a single implementation

• Market Adaptation: Different variations may perform better in trending vs ranging markets, providing natural portfolio balance. I hope to get a better understanding for this with the use of an AI driven trading analyst I’m building.

This framework has proven especially valuable during various market conditions throughout 2024.

And we're starting 2025 with a bang—$27K profit in the first week of January. As they say in trading, past performance doesn't guarantee future results, but it sure beats the alternative. Scroll to the bottom for a screenshot of last week’s performance.

Important Announcements

1. Portfolio Strategy Changes

The seven strategies in the Q1 2025 Forward Test represent our cream of the crop. While we haven't tested all 90 published strategies, these are the current champions. These strategies will be removed from the ATS paid subscription on January 30. Don't worry—I'll continue sharing performance updates in the Mudder Report, but access to these specific strategies will be available only through a separate package. Please take a moment to download the Forward Test strategies before January ends.

For links to all strategies click here

2. New Pricing Structure

In response to community feedback, I’m adjusting pricing. Starting March 1, 2025, the paid subscription will be reduced to $255/month. (Annual subscription rates remain unchanged.)

3. Introducing Strategy 83c

Strategy 83 has been one of our crown jewels, created using high-performing entry and exit commands based on MAE, MFE, and ETD statistics. Its star variant, Strategy 83c (previously unpublished), has 8 variations in the Q1 2025 Forward Test and generated:

• Q4 2024: $55K profit

• First week of Q1 2025: $13K profit (almost half of total profits for the week)

Rather than auction this strategy, I’m offering it as a special reward to loyal subscribers. To receive Strategy 83c:

• You must be a paid subscriber for at least 30 days

• The last day to subscribe to receive Strategy83c is January 29, 2025

• Strategy83c will be emailed (not posted) directly to qualifying subscribers on February 28

Looking Ahead: The Next Frontier

As we celebrate exceeding our original benchmarks, and because nothing in trading is written, we're setting our sights on new horizons as well. Stay tuned for these upcoming dates for Q1:

Key Posts and Dates

• January 14 - Post: The New Paradigm: Cash Flow Trading

• January 20 - Podcast 1 - Using Agentic Frameworks in Automated Trading Part 1

• January 25 - Post: Complex Systems Theory In Trading Strategy Behavior

• January 27 - Podcast 2 - Using Agentic Frameworks in Automated Trading Part 2

• January 29 - Strategy 83c cutoff date

• January 30 - Removing strategies in the Q1 2025 Portfolio from the paid ATS subscription plan. Strategies and updated parameters will be available as a package on March 1 for $10,149.

• February 3 - Podcast 3 - Using Agentic Frameworks in Automated Trading Part 3

• February 11 - Post: A Framework for LLM Based Automated Trading Information Systems

• February 19 - Post: The Impact of A Digital Workforce On Assets & The Cost of Labor

• February 28 - Email: Sending Strategy 83c to all qualifying subscribers

• March 1 - Lowering the price of a paid subscription to $239/month. There will be no change to the annual subscription.

Additional Q1 2025 Content

• Regular updates on Q1 2025 Forward Test

• Strategies 89 & 90 progress (including ML/continuous learning)

• Frequency optimization insights

• Agentic platform developments

• Incubator portfolio (targeted mid-February)

• A special announcement planned for March 5

Trading is a journey, not a destination. Here's to making 2025 our best year yet!

If you have any questions, start with the FAQs and if you still have questions, feel free to reach out directly at AutomatedTradingStrategies@protonmail.com.

See below for a link to the Q1 2025 Forward Test Portfolio as well as an overview of Week 1 performance results.

Read more

I hope everyone had a great summer. I had some truly inspiring conversations, listened to some great live music, doubled my bird count, and got lots of sun on the beach, but now I’m back in the office and ready for what promises to be an exciting quarter. I hope you are too.

Okay, let’s get started.

Due to a number of cyclical patterns that you can read about here, I knew 2024 was going to be a wild ride. I’ll provide a quick NQ technical update at the end of this post to show where I think the market’s going for the rest of 2024. I’ll also provide you with a first draft of the ATS Q4 Portfolio, which is the best yet. I know I say that every update, but this is truly a special quarter. I’m also going to provide a quick update on where we are with the new system build, which is based on the efficiency of communication with the most accurate and capable LLM.

Why am I structuring our quest around an LLM? There can be only one reason. I believe it can help us to create the holy grail of automated trading strategy. In my opinion, there is no better team member.

I want to take a few minutes to revisit what Ray Kurzweil said about LLM’s during a Joe Rogan interview in March of this year. He said that LLMs should actually be thought of as much more. Instead of large language models, Kurzweil said we should be calling them large event models, because “they do all kinds of things”. You can listen to the conversation yourself below starting at ~21 minutes below.

When I asked the LLM Claude to provide a more descriptive ‘term’ for its technology, it said that the best term would depend on which capabilities the AI is being used for. Fair enough.

Kurzweil intends to use LLMs to achieve immortality. I intend to use it to find the holy grail of automated trade strategy. The challenge with LLMs is accuracy, but with the help of techniques like reflecting and mirroring, the gap is quickly closing.

AI Sapience: "Fake it till you make it"

Reflection and mirroring in the context of AI and LLMs are related, but distinct concepts. Reflection is about AI's ability to analyze and think about its own processes and outputs, while mirroring refers to AI's ability to emulate or copy specific behaviors, styles, or patterns. This could be in language use or problem-solving approaches.

Just as a chef might adjust a recipe based on previous outcomes and feedback, a reflecting prompt may allow an LLM to fine-tune its trading parameters based on past results. Likewise, similar to how a skilled dancer mirrors their partner's movements, a mirroring prompt may allow an LLM to synchronize its actions with the rhythm of the market. Put together, these two algorithms make a powerful ensemble, especially as a prediction model.

Read more

Read more

FAQs

Housekeeping:

My apologies for getting this out later than expected. I’ve been traveling and there’s bad weather in the area.

Also, a few of you have asked me if I’m trading any strategies live. I understand the confusion. Hopefully, I’ll have an opportunity to create a timeline in the fall. As a quick review, I used to only publish backtests. Backtests are third fiddle to forward tests, and forward tests are second fiddle to live tests.

Backtests are run on historical data using a simulation. This allows for errors in the historical data file as well as in the way the data file is used (OHLC) to simulate bars. Bars that require more than these four data points are often wrong. Bars that take longer to form (60 minute) are more reliable than bars that form quickly (1 minute). In other words, you can’t rely on backtests: garbage in-garbage out.

Forward tests use live data on a simulated account. So the data is real, not historical. The chart is also live, so the simulation can use more than four data points to create each bar. As a result, I can confidently say that the performance data received from forward tests is better for decision making. That doesn’t mean that the strategy is going to behave the same in the future. It merely gives us a better understanding for how an animal moves.

Knowing what I know now, I would never run a strategy live that I hadn’t forward tested for at least a few months first.

Live Tests are the best, but they are also the riskiest. Unlike the ATS Forward Test, the ATS Live Test was conducted on a live account, not a simulated account. ATS ran our first live test in January of 2024. It ended in June up 196%. It was terrifying at first, but then the process of “letting go” got easier. The experience provided many insights that you can read about in Live Test updates. Needless to say, my goal for the next four months is to find the best strategy or strategies to run in the next Live Test.

I hope this helps those of you that may be new to automated trading and/or ATS. Please let me know if you have any additional questions and click here for a list of FAQs.

Let’s get into it…

What a week guys! After last Monday’s panic selling, the market nearly wiped out its losses for the week. The performance played out in the ATS Forward Test portfolio as well— net profit came in flat to positive at $1,428 on 671 trades.

It all started two weeks ago, actually. After the Fed announced no change, traders that had baked in easing threw a tantrum and markets panicked. If nothing else, we learned that the market is extremely sensitive (and far-reaching) these days. From investors and business owners crying over no-rate cuts to an unwinding of the yen carry trade, the markets are finding it increasingly hard to settle on price. In other words, there is a great deal of uncertainty in the air. Looking at it from a trading perspective, this is part of a pattern: a cooling-off period before the next run-up. Don’t forget, GDP is still at 2.8%.

There are those that say volatility is the price of return, put another way, risk is the price of profit, but I’m convinced that does not have to be the case. After running a live test from January to June of this year using ATS strategies, experiencing all the ups and downs, and the emotions that come along with those ups and down, I’ve changed my fitness test a bit. Instead of a focus on one factor, ie, net profit or profit factor, my fitness test for strategy viability is a combination of three things: max adverse excursion (MAE), profit factor and trade count. Net profit, is still the primary filter, but these are the metrics I use to prioritize. Prioritization is critical when trading multiple strategies, especially if they are highly correlated.

Two weeks ago we looked at:

• which strategy variations performed well

• which instruments performed well

• how well AI-generated strategies performed ag. non-AI generated strategies

This week, I’m going to provide you with three different lists:

• a list prioritized by net profit

• a list prioritized by profit factor

• a list prioritized by MAE

Depending on your goals and account type/size, each of these lists can help to align and prioritize your portfolio with your goals.

We know that almost all strategies in the portfolio perform well when the market goes up. Some do better than others. Likewise, some do better than others when the market goes down. With such a volatile and crazy market, the question is: which strategies were able to do both?

Strategy Links

Well, out of the 140+ strategy variations in the Q3 Forward Test, these are the high performers. It’s important to note that the strategies in the Q3 Forward Test were chosen based on their overall performance in the Q1 and Q2 portfolio, which is to say, their performance has already been confirmed by forward tests on some level. We’re getting close…

Top Strategies By Net Profit

Here’s a chart of the top 20 strategies in the Q3 Forward Test so far by net profit…

Read more