Why do trading strategies stop working? When do you stop trading? How can you avoid or minimize that trading strategies stop working? These are the most important aspects in trading because most trading strategies stop working – sooner or later. Many traders would hope a strategy lasts forever, but markets are not meant to be static. Markets are dynamic and evolve and change both gradually or suddenly. You better be prepared or at least minimize damage if (or when) it happens.
Strategies stop working mainly because of curve fitting, structural and cyclical changes, survivorship bias, behavioral mistakes, commissions, and slippage. Short-term trading is a zero-sum game and you need to accept that trading strategies at one point stop working. You better be prepared!
The better the foundation of your business plan, the fewer trading strategies stop working. The more you prepare for inevitable trading hiccups, the less they matter. Imagine yourself having one or several trading strategies that are literally handing you money on a silver plate. Then one day everything stops working and you are stranded with no strategies that seem to work! You don’t want to be in that situation.
Unfortunately, trading strategies do stop working. In this article, we look at reasons why trading strategies are not working or stop working and how you know a trading strategy stops working.
We backtest trading strategies and systems on a daily basis. If you are looking for a short term trading system, please click on the link (we have made hundreds of strategy backtests).
Trading is all about backtesting and generating ideas
Markets change all the time. Good traders know this and adapt by constantly looking for ideas to put into live-action. By being prepared you minimize the damage when strategies stop working. We have personally had very profitable trading strategies closed literally from one day to the next due to new regulation and legislation. You never know what is being thrown at you.
Any quant trader should spend at least 80% of their time backtesting and brainstorm for ideas. There are plenty of ideas both for free and as subscription services on the internet. This website has hundreds of free trading systems and strategies, and around 100 paid premium strategies for our members.
What are the chances of finding a tradable trading strategy? Probably around 1 in 20 backtests show promise. Of the one in 20 that passes our backtesting criteria, very few make it past the incubation period (see below).
As you can understand, this is a time-consuming process! But in the long run, you get rewarded because you’ll have fewer strategies that stop working and your trading becomes more robust.
Understand why strategies are not working
To better understand why a trading strategy is not working (or perhaps stops working), you need to understand why trading strategies fail:
Not long enough backtest and no out of sample test
A backtest needs to generate many trades to be of any significance. Moreover, it needs to be of significant length. A lot of traders only use a limited time frame and thus are more liable to randomness and cyclical trends in the market. Make sure you have tested your strategy over many years and in many different investment climates.
For example, the bull market from 2008/09 has been pretty long and all driven by quantitative easing from central banks. Prior to 2008/09, there was no quantitative easing. Thus, the momentum strategies that worked then will most likely face a pretty hard time when the easing stops.
Put simply, most trading strategies are not adequately backtested. Some strategies work fantastic a couple of years before they fade away. Breakouts might work in a rising market, but less so in a sideways and falling market. Be sure to test in all types of markets.
And make sure you have a proper out of sample test. Divide your dataset into two parts: one for in-sample and one for out-of-sample. Read more here:
Also, an incubation period of many months can save you a lot of money:
Incubation period: test every trading strategy live in a demo account
When you have found a strategy that looks promising in a backtest, you should paper trade it for many months in a demo account.
We like to call this the incubation period. It’s time-consuming but this “trick” saves you money in the long run and makes it less likely your strategies stop working. And when your strategy stops working, you are prepared for it.
If you have done a live test and are pleased with the result, you can start trading it live. The majority of the trading strategies we test via incubation never make it to live trading with real money.
The strategies that pass the incubation period normally last for many years. The incubation period is an excellent way of minimizing the strategies that stop working.
Always quantify your ideas
We believe automation is superior to discretionary trading. Automation has two main advantages.
First, you can trade and execute an unlimited amount of strategies. The computer does all the trading for you.
Second, you minimize second-guessing and behavioral mistakes. The less you stare at the screen, the less likely you are to screw up by doing something not planned. The fewer screw-ups, the less likely a strategy stops working.
Why? Because many strategies stop working because you start tweaking your strategy or skipping signals. Automation removes many behavioral mistakes.
Trading strategies stop working because of curve fitting
When something is curve fitted, it is just a question about time before it ends up useless. Curve fitting happens because of too many parameters and criteria and too short a time period for the backtest.
The best trading strategies for the long term are those which have the fewest parameters:
Simplicity reduces strategies stop working
Many traders spend months tweaking one “super strategy”. We believe this is wrong for two reasons:
First, you risk curve fitting or going round in circles. You should be very careful in changing parameters, at least only after rigid backtesting and incubation.
Second, you risk ending up with more parameters than necessary (and thus curve fitting). The best trading strategies are always those which have the fewest parameters.
Third, you want many strategies, not just one good one. Several “not so good strategies” are highly likely much better than a “super strategy”.
Structural change makes trading strategies obsolete and they stop working
In our opinion, the best trading strategies are those that are based on some structural edges.
By structural edges, we mean an edge that is built on how an exchange operates, for example. The specialist system on the NYSE was another example of what we consider a structural edge.
We had great success for some years trading the opening imbalances at the open on NYSE, a strategy called opening price trading strategy. But all good things come to an end, and the change to electronic trading made the specialist system almost obsolete. Thus, our best strategies simply stopped working.
How the stock exchanges operate has a tremendous impact on how strategies work (or do not work).
Another example of a structural market edge is the turn of the month effect (this one could be labeled cyclical as well). We suspect this effect happens because investors and savers allocate more money to stocks at the end and the beginning of each month.
We believe it’s easier to diagnose any problems with structural effects than many other edges and they are less likely to stop working.
Cyclical change makes trading strategies difficult to trade and follow
Some strategies work well in certain types of markets and not so well in others.
For example, trend following has always had many years of underperformance before they start working again. This makes them very hard to follow and trade, and that’s perhaps the reason why they seem to work in the long term.
The Dogs of the Dow strategy was once a very good strategy but less so the last ten years. Is this cyclical or a structural change? Permanent or temporary? We don’t know but the markets have evolved and changed since the strategy became very popular.
For example, Ben Bernanke started quantitative easing and this has had a huge impact on asset prices and behavior. Is this permanent or temporary? We don’t know, but it sure changed the market!
Being dependent on one type of strategy makes you vulnerable
Some only trade mean reversion, and others only trade trend following. Likewise, some are day traders and some are swing traders.
We believe this is a mistake for many aspiring traders. You should be agnostic and trade anything that works. You need to diversify as much as possible to smooth earnings.
One dollar made in crude oil is the same as one dollar earned in day trading Microsoft. One dollar made day trading is the same as one dollar made in a weekly time frame. We believe the most rational approach is to diversify to different time frames. Why? Because not all time frames stop working at the same time.
- Uncorrelated assets and strategies – benefits and advantages
- Does your trading strategy complement your portfolio of strategies?
Strategies are not working because of survivorship bias
This is something all traders ignore (or forget). They tend “forget” because survivorship bias should be just a minor problem – a detail?
No, unfortunately, survivorship bias can have a huge impact. We have covered this in a separate article which we strongly recommend reading:
Behavioral mistakes make you not follow the signals of the strategy – thus you have no strategy
The first requirement to a trading strategy is that you should trade all signals the strategy tells you to do. But trading biases always put a spanner in the works. It’s easy to skip a trade after four losses in a row or if you are in a drawdown!
But this means, in reality, that you don’t have any strategy in the first place if you skip trades. It’s almost impossible to know before you trade a signal if it’s going to be a winner or loser. The markets are unpredictable and usually not intuitive. You have not backtested omitted trades, and thus you have no strategy. Many quants become discretionary traders by skipping trades.
Commissions and slippage are underestimated
Because of structural changes, like described above, slippage might increase or decrease depending on certain factors. For day traders this might be the difference between a lot and nothing.
However, if you stick to very liquid assets, slippage should not be a problem. We measured slippage in live trading.
Inefficiencies get arbed away
Eventually, all inefficiencies in the markets get arbed away. A strategy can become too well known, for example, when a book is written about the strategy.
Short-term trading is a zero-sum game
Always keep in the back of your head that short-term trading is a zero-sum game. If you invest for the long term, you get a tailwind from the gradual increase in prices (inflation) and earnings growth. This takes time to get reflected in increased share prices, but traders don’t have this luxury.
The options and futures markets are a 100% zero-sum market – even negative considering the costs. What you make, someone else must lose.
Because trading is a zero-sum game, you can’t expect trading strategies to work forever. They sooner or later stop working:
Almost all trading strategies stop working – sooner or later
If you find a good trading strategy you need to accept that sooner or later it will stop working – preferably later, of course.
This is the sad fact of trading. If you’re a long-time buy and hold investor, you don’t need to worry about this. But if you’re a trader, you need to understand that trading is a constant battle of having an arsenal of trading strategies.
How do you know a strategy stops working?
If you know that all trading strategies sooner or later stop working, you are somewhat prepared. If you do the following, we believe you are better prepared and minimizes the risk of strategies stop working:
Look for abnormalities
Abnormalities normally revert to the mean, but of course, not always. Trading is a numbers game.
Stock prices are not normally distributed, sometimes we get a black swan, but we believe mean reversion is the lowest hanging fruit in trading. Jim Simons also agrees that mean reversion is low hanging fruit.
Correct position sizing
Setting together a portfolio of trading strategies is not easy. The thing that complicates the most is the position sizing.
For example, if you trade 5 different strategies in stock indices, do you just trade one signal at a time (not taking a signal when you already have a position), or do you still trade the signal to increase your overall exposure?
Are you tempted to increase size after a period with good results? There are many temptations to adjust the position size along the way, and often it will be the wrong thing to do.
Wrong position sizing makes you believe a trading strategy has stopped working, but in reality, many times it all boils down to the correct size of your positions.
Our best advice on position size is to trade smaller than you like:
Prepare for inevitable drawdowns by trading small
Always trade smaller than you like. If you have a good strategy or a good period, it’s very tempting to increase size in the belief that you’ll make more money. Many want to get rich in a hurry! We believe it’s better to get rich slowly by being patient and let your capital compound.
But it’s not that easy. If you get in over your head, you make inevitable mistakes, especially behavioral mistakes. Many increase size after a good period, only to lose more when the inevitable drawdown happens. During the drawdown, you reduce size….. This is a vicious cycle you want to avoid.
The best medicine for avoiding behavioral mistakes is to always trade a little smaller than you would like. If you want to get rich in a hurry, you increase the risk for a substantial setback sooner or later when strategies stop working.
Make sure you use the Holy Grail of trading – diversify
The Holy Grail of trading is having many different trading strategies that are uncorrelated to each other. That is, of course, difficult, but it should be your goal.
When you can trade an unlimited number of strategies, only your imagination and idea generation stop you from trading plenty of strategies. You should diversify on markets, time frames, and types of trading strategies.
This is a topic we have covered before in these articles:
The point with diversification is to have strategies offset each other but still make a positive return.
Thus, when your strategy stops working, you are prepared. It’s not likely that all your strategies fail you at the same time if you are properly diversified.
Why trading strategies stop working and how to avoid it – conclusion
The truth is, perhaps sadly, that the future is unknown. On the other hand, this is what makes life worth living. If all was plain sailing it would be rather dull and boring (?). You can’t eliminate the risk that your trading strategies stop working, but you can prepare for it and minimize the risk.
What do you do to prepare for the time when your trading strategies stop working? Do you panic and turn off the strategy? Or do you just keep on plugging as if nothing happened?
The main problem in trading is that there is no way for sure you can tell if a strategy is finished or just in the middle of a normal and expected drawdown.
However, if you have a clear idea of why the strategy is working in the first place, or at least why it should work, you can better tell if you should continue or stop trading the strategy.
Moreover, you avoid or minimize trading strategies stop working by making proper backtests with many months of incubation testing in a demo account, you diversify into strategies by trading different assets and time frames, and you look for trading edges where you can have a sustainable edge.
The markets are non-stationary and adaptive, and thus most trading strategies stop working (sooner or later).
