Automated Trading Strategy #76: Irrational Cycles & Patterns of Human Behavior
Part II (Update)
Important: There is no guarantee that these strategies will have the same performance in the future. I use backtests to compare historical strategy performance. Backtests are based on historical data, not real-time data so the results shared are hypothetical, not real. There are no guarantees that this performance will continue in the future. Trading futures is extremely risky. If you trade futures live, be prepared to lose your entire account. I recommend using these strategies in simulated trading until you/we find the holy grail of trade strategy.
Update: Material changes have been made to download instructions and Ninjatrader Strategy Builder screenshots.
This is Part II of a two part post introducing Strategy 76.
In Part I, we looked at the type of trading behavior the quant fund Renaissance Technologies looks for to find opportunities in the market.
In Part II, we’ll look at what causes that behavior, why it happens and how to spot it in yourself to improve your trading (both manual and automated). We’ll also look at how to use what we learned from Renaissance to create Strategy 76.
Let’s get started…
Cycles & Patterns of Human Behavior: Are You Prioritizing Certainty Over Price?
In Strategy 76 Part I, we learned about the strategy behind the quant fund Renaissance Technologies. We learned that the ‘secret sauce’ comes down to identifying irrational cycles and patterns in human behavior. One way to study these irrational cycles is by observing the bidding behavior in a Dutch auction.
There’s a surprising degree of literature on the subject of Dutch auctions. Most papers are geared toward coming up with ways to optimize price, not profit. In other words, most are from the perspective of the broker or seller, not the bidder. Conclusions about the nature of the bidder’s internal reward system like the ones found in the journal paper: Going, going, gone: competitive decision-making in Dutch auctions are rare and give a great deal of insight into the behaviors that drive markets. The paper was published by Cognitive Research: Principles and Implications and is a collaboration between universities located in both Australia and the United States.
What is a Dutch Auction? According to the paper:
A Dutch auction is a descending price auction where an item begins at a set maximum price. The price is gradually lowered over a fixed period of time until a bid is placed that guarantees the bidder the purchase of the item at the current price (Thomas 2012).
The NQ futures market is essentially two Dutch auctions (buy/sell) working against each other to find a ‘price’ so studies on Dutch auctions are a great place to gain insights on trading behavior.
Why is that?
Much like the market for standard securities, a Dutch auction creates a cascade of decision making that has been studied to the point of exhaustion by those in the business of making markets.
What is the decision framework of a Dutch auction? Here’s an explanation from the paper.
Decision-making behaviour in a Dutch auction requires participants to balance the speed of response (i.e. time of bid) with the consequence of that response (i.e. price). Bidding earlier at a price equal to or greater than the value of the item increases the probability of winning but also of overpaying for the item, which would result in no resale profit (Vickrey 1961). On the other hand, the longer a bidder delays in placing a bid the higher the likelihood the bidder increases their profit on resale, but the risk of losing to a competitor also increases. The way in which a bidder trades-off between the certainty of winning and the price they pay is the key to developing a bidding strategy within a Dutch auction (Easley et al. 2010) and makes for an ideal context for the study of competitive decision making.
To be clear, you are the bidder. In games of probability, the primary reason for failure is poor decision making, which leads to more poor decision making, which leads to more poor decision making. Based on the literary review above, the longer the bidder delays in placing a bid, the higher the likelihood the bidder increases their profit. This is an important point, so I’m going to say it again: the longer the bidder delays in placing a bid, the higher the likelihood the bidder increases their profit. This is one reason I like to study the market for at least an hour before making a trade. If possible, I will also conduct a market replay of the night session.
So there you have it. That’s the answer. The best way to increase your profitability is to simply wait for the perfect trade to come along. We all know that’s easier said than done. It is the cognitive struggle that all traders experience. Whether they know it or not, fears exact a toll in the form of poor decision making. Some people are willing to pay a premium to reduce the effect of this fear. It’s important to be clear about this: this person is now prioritizing the reduction of fears over making a profit. That is, there is a known trade-off that human beings make between certainty and value, even if that certainty increases the probability of a bad trade. Here’s another excerpt from the paper:
To gain certainty in the context of a Dutch auction, a winning bid must be placed earlier and at a higher price. Ample evidence suggests people prefer certainty over uncertainty in a variety of conditions (e.g. Kahneman and Tversky 1979, 2013).
Again, this is telling us that there are people that (whether knowingly or not) would rather pull the trigger now than wait for the better price. In other words, some traders will overbid to avoid the emotional state caused by missing out on a trade, which only increases the odds of a loss.
Do you recognize this behavior in yourself? I do. It happens to all of us. I once sat next to a trader that made the highest bid for EUR ever marked. He said he couldn’t help it.
Instead of scaling into a position to reduce risk, you are scaling to assuage your fear of missing out. Do you feel worse when you lose money on a trade or a trade that you waited for and lost money on? If so, are you willing to pay money to even out that pain? Do you fear the loss of profit or the missed opportunity more? If so, you are not alone. According to the paper:
Certainty matters; bidders may prioritise certainty of winning in an attempt to avoid the negative emotional state caused by losing. Adam et al. (2012) found participants felt the loss of a Dutch auction more strongly than a win, and suggested participants were better able to cognitively prepare for the win, as, in placing a bid, they expected certainty in obtaining the item. A loss, on the other hand, was unexpected causing a greater emotional response.
In a nutshell, some of us are more concerned with certainty than profit. Some traders will even neglect optimal bidding strategy to save time.
If a bidder prioritises time, they may place an earlier bid. This would result in a higher price but at the same time (pardon the pun) also enables bidders to save time in auction participation. Lucking-Reiley (1999) proposed that bidders may neglect optimal bidding strategy and save time by terminating auctions prematurely
Wow. That’s a hard circle to square. But, I can see these behaviors in my own trading. I take manual trades personally — it’s me against the market, whereas automated trading is a system against a system. There’s less emotion. This is due to a perceived notion of lowered competition. According to the paper:
With the increased certainty of winning in Dutch auctions, there also comes an increased risk of an immediate and certain loss to another competitor. This risk may have a range of effects on how bidders determine the trade-off between certainty and price. Ku et al. (2005) proposed that the level of perceived competition could result in an intense emotional state of competitive arousal which reduces an individual’s ability to think clearly and increases risk-taking behaviours.
This is akin to what poker players refer to as “tilt”. Poker players have a similar decision framework that can impact the player’s relationship with certainty and price. Trading with an automated system can help to prioritize profit over certainty thereby eliminating the risk of coming in at the top of an upward trend or the bottom of a downward trend, aka, the winner’s curse.
Prioritising certainty helps bidders avoid negative emotions associated with losing, but also increases the risk of falling victim to the winner’s curse. The winner’s curse is the potential for the winning bid to exceed the value of the item being auctioned (Easley et al. 2010).
This leads the bidder with the highest estimation of the item’s true value to win the auction for an exorbitant value (hence the term ‘winner’s curse’) (Van Den Bos et al. 2008).
Why is this so important? Automated and semi-automated systems can help you navigate and hack the market.
Hacking The Market
The market is a kind of psychological operation. We talk about sharks and whales as types of traders, but these are also ways to think about the “type” of psychological attack on your emotional state. The goal is to knock you off your game; to make you buy at the top and sell at the bottom.
The best way to fight any PSYOP is to recognize that you’re in one. What role are you playing and what role do you want to play? Are you a dentist or trader? Figure out a way, either through a trading group, mentor, journaling, etc, to get feedback on your trading patterns. According to the paper, obtaining greater awareness about your trading patterns is the only way to identify them and change the behavior:
Garvin and Kagel (1994) found that bidders adapt their bidding strategies to reduce the winner’s curse. They showed that while inexperienced bidders initially had high rates of the winner’s curse, these bidders could learn to reduce the effect through feedback. Furthermore, they found this learning process was aided through the observations of the winner’s curse in other bidders. This suggests that an individual’s bidding behaviour is a dynamic and malleable process that can be continually adjusted through their bidding experience and through their opponents bidding experience.
However, while this study showed that bidders can adjust their bidding strategy through learning, the effect may not be as influential as first thought. Indeed, Lind and Plott (1991) found the winner’s curse was still present in experienced bidders albeit at a reduced magnitude.
Bottom line—with practice and feedback, the novice can become the pro. Even experienced traders have blind spots, which is where automated trading comes in. Automated trading eliminates the “uncertainty” that causes traders to make poor decisions including:
the fear of missing out,
the fear of waiting for a bad trade,
the fear of wasting time,
the fear of the negative emotions that come along with a loss or missed trade; and,
the fear of increased competition.
Each one of these can have an associated ‘cost’ if not managed. To that end, Strategy 76 and 77 are based on an indicator that many experienced traders use, but for whatever reason I didn’t think to use it myself until after reading the paper referenced above. Unfortunately, it isn’t included in the native indicator suite for Ninjatrader. Thankfully, Larry Kann, writer of Hunt Gather Trade, was able to create the indicator for us.
These are the performance results for Strategy 76:
This is a portfolio of 10 equity future contracts (1 contract per instrument). Six of the instruments are micro contracts and four are mini contracts. In total, the model shows that this strategy makes 172 trades throughout the year, has a percentage of profitable trades of 70% and a net profit of $59K. It is important to note that the percentage of profitable trades is organic, not forced. It is also important to note that this represents only one side of the trade. I want to build on this strategy so I think it’s important to start with one side.
The beauty of this strategy is that it can be applied across instruments in a way that is highly customizable yet hard to over-fit. And, Strategy 77 will take Strategy 76 to the next level with the use of a forecasting model. I’m getting ahead of myself. First, let’s talk about how to recreate Strategy 76.