Drawdown is one of the most important topics to understand and master when you learn stock trading. It is given a lot of lip service in trading system development literature but there is very little practical information on how to manage and reduce drawdown in your stock trading systems.
Most stock traders generally understand that you don’t want to have large drawdowns and you want to limit your risk, but realistically, most traders focus on return when they are developing trading systems. In this post I am focusing on the drawdown side to bring up some of the issues and share some interesting issues and ideas for you to consider in your trading system development.
I first want to give a clear definition of drawdown in stock trading and trading systems, just to make sure we are working from the same baseline. When you run a back test, you get an equity curve that shows you how your portfolio would have grown over time if you had followed those trading rules in the past. At any point on the back test, it might make a new high relative to the previous equity values in the back test. As you move forward from that new high, it could keep going up, in which case there’s no drawdown or the equity could fall, in which case that’s a drawdown.
At any point on your equity curve, you’re looking forward to see what the largest percentage decline is from a equity high to a subsequent equity low. This percentage decline is the drawdown that your trading system experienced from that point. In Amibroker when you run your backtest you get two main graphical outputs.
The first is the equity curve which looks like this:
The second is the underwater equity curve chart. This shows the size of drawdowns at each point on your equity curve, how frequently they occur and how regular they are.
The underwater equity curve you get when you backtest your trading system in AmiBroker shows you the percentage drop from the highest point on your equity curve to the subsequent low. This shows you what the percentage drop in your equity would have been had you been trading the system at that time.
You should be aiming to minimize the drawdown in your trading as much as possible for a given level of return… or maximising your returns for a given level of drawdown. There is always a tradeoff between drawdown and return that we have to work with as systematic traders.
The important difference between trade drawdown and portfolio drawdown
So there’s one thing which is often misunderstood when new traders think about drawdown, and that is the difference between trade drawdown and portfolio drawdown. Many traders will freak out or get concerned when they have a trade and it dips 30% or 40% before exiting or before going on to new highs. They have already planned on a drawdown tolerance of say 20%, but here they are with a trade drawdown of 40%. The conversation usually goes something like
“I was only planning to have a 20% drawdown, but I’ve got this trade with a 40% drawdown so something must be broken, something has gone wrong, I better liquidate this trade”
There is actually nothing wrong because on a trade basis, the drawdown can be very different than on a portfolio basis where you’ve got many trades compared to an individual trade level. When you backtest your stock trading system, the portfolio level drawdown is the one that you need to be consistent with your psychology and your risk tolerance.
The trade level drawdown is quite different to the portfolio drawdown and is worth understanding and investigating because you can get nervous on an individual trade. Your ability to follow your system can be sabotaged by an individual trade. Even if that individual trade is within the bounds of what’s normal for the system and even if you’re not in a bad drawdown at a system level, one trade can still cause you concern.
When you are looking at individual trade drawdown you use the maximum adverse excursion from the entry of your trade to understand how far each trade can move against you according to your backtest. The Maximum Adverse Excursion (MAE) on a trade can be much higher than your portfolio level drawdown tolerance because many trades combine together to form your portfolio, so they tend to smooth each other out in the portfolio.
So please do not get concerned about individual trade MAE if the system is achieving your objectives in terms of return and drawdown at a portfolio level. You can use Maximum Adverse Excursion data to find ways of improving the accuracy of the entry signals in your trading system, but it certainly should not be used in any discretionary way to close a trade that has not generated a sell signal on its own.
Why you need to manage your drawdowns and keep them small when trading stocks
So why do you need to keep your drawdowns small? There are many reasons including
- The trading psychology importance of managing drawdowns
- Ability to recover to new equity highs
- Impact on compounding & Interaction with your other systems
- Blowing up your account and starting again
Each of these is covered in the sections below…
The trading psychology importance of managing drawdowns
So why do you need to keep drawdowns small when you are trading stocks, or any other instrument for that matter?
The first and most important consideration is your trading psychology. It’s your money, and you want to do something cool with it. You want to buy a new house / car / freedom etc with it. Whenever your account drops, that trading goal is threatened, whether you like it or not. The trouble is, your trading goal is threatened in direct proportion to the size of your drawdown!
The trouble we then have is that our animal instincts kick in – whenever we feel threatened we move into “Fight or Flight” mode. It is a survival instinct that is hard to avoid (something to do with a few million years of evolution – The Fight or Flight response is explained here) When we feel threatened as a species, humans stress levels go up and our blood is redirected to our muscles to get us ready to either fight for survival or run (Thus the name ‘Fight or Flight’) the problem with this is as soon as this happens our intelligence falls because our body is ready for action… it is no longer ready to make rational and sometimes complex financial decisions.
The bigger the size of the drawdown, the more we feel threatened and the more stress we feel, so the lower the intelligence and the worse financial decisions we are likely to make.
The financial markets are the biggest competition in the world, and if we are going to survive we need to be at the top of our game. Being overwhelmed with stress because of a large drawdown will ensure we are not at the top of our game… and we are more likely to be carried off the field on a stretcher after blowing up our account.
The key to managing your trading psychology when it comes to drawdown is to understand our drawdown tolerance. In The Trader Success System I give my students a simulation which they use to determine the level of drawdown they start feeling uncomfortable. If you have been trading for a while then you probably have a good idea of the percentage drawdown that caused you to get uncomfortable previously. If you are new to trading then you need to visualise your account falling from current levels and imagine how you would feel at different levels of drawdown. As soon as you start to get uncomfortable, that is your tolerance level.
Once you know your drawdown tolerance level you need to ensure that your portfolio drawdown remains below that level, so you can continue to trade without being emotionally compromised.
So we need to be very mindful of the size of the drawdown we’re currently in and our tolerance levels, and make sure that we are not letting ourselves panic when we’re approaching our tolerance level. We have to observe that we’re getting nervous and journal about these thoughts because whenever those psychological stresses occur, the pressure we feel builds up and we risk making poor trading decisions as a result of the stress we are under.
When we are at or over our drawdown tolerance, all it takes is one more loss all of a sudden you could find yourself freaking out and then all of a sudden it explodes into non systematic trading behaviour and you lose even more money!
Ability to recover to new equity highs
Second major reason you need to control your drawdowns and ensure they are small is your ability to recover to new equity highs. The asymmetry of drawdown recovery is one of the most challenging aspects of trading.
If you have a 10% drawdown, you have to make 11% on your equity to get back. If you have a 50% drawdown, you have to make 100% on the remaining equity to get back to break even! For even larger declines it gets even more out of hand as the image below shows.
I don’t imagine recovery from a 90% drawdown is a very likely proposition!
The next question you should ask yourself about drawdown recovery is “how long will that take?”
Do have the mental fortitude to keep trading after a huge drawdown?
If you were up at a million dollars and you’re now down at $500,000, and you’re looking up at that hill, trying to make it back to a million, you’re probably feeling completely overwhelmed and you are not in a good space to trade well. Whether it’s $1,000 or $5,000 or a $1,000,000 the psychology is the same. It’s all relative to where you’ve been, your experience, the elation of the equity high you felt previously and the level of overwhelm you feel now at your ability to claw your way back.
Your ability to recover psychologically from a drawdown and keep trading is critical!
Impact on compounding & Interaction with your other trading systems
The third point is the impact of the maximum drawdown (MDD) on your longterm compounding. When you run a back test with one system, you might say, “Oh, it’s got a 50% drawdown, but man, this gives me the best longterm return.” That’s great in isolation.
The difficulty comes when you have multiple systems, they feed off each other because they’re drawing from the same equity pool. So if you have one system that has a 50% drawdown, it stunts the position size and growth in your other systems, and as a result it stunts the compounding in your other systems too. So even though they may be offsetting each other, a system with a huge max drawdown potential could really be doing a disservice to your portfolio as a whole.
So when you’re combining multiple systems, often it can be better to actually reduce the risk and focus on reducing the drawdown so that they can feed off each other and all compound on top of each other’s gains more quickly, rather than having one big draw downing system.
I much prefer managing the drawdown by designing each of my trading systems to keep the drawdown small so that the whole portfolio of systems work well together and I can achieve smoother portfolio growth.
If you have a portfolio of non-correlated trading systems that is fantastic, but if you have a portfolio of non-correlated trading systems with low drawdown, that’s even better!
Blowing up your account and starting again
The final reason you should aim to keep your maximum drawdown small is to minimise the risk of blowing up your account. While your backtest gives you a lot of information about the expected system drawdown, it is not perfect because the max drawdown in real time trading could exceed your backtested trading drawdown (This is explained further in the next section). If you have a large drawdown over 60-70% then your ability to recover is close to zero because you need a HUGE return to get back to new equity highs.
Unfortunately when traders find themselves in a large drawdown like this there is a huge temptation to trade aggressively to make back the losses. This trading behaviour will almost always end in disaster and blowing up your account. If this happens, the very unfortunate thing is you will end up either quitting trading or have to spend time to save the money to start again. Look what happens to your future net worth if you blow up and lose your account and have to start again:
Maintaining a low drawdown and preventing your account from blowing up is not just about the small amount of money in your account now. The impact on your future wealth is devastatingly large, so avoiding huge drawdowns and blowing up your account is critical if you want to become a wealthy and successful stock trader!
How to determine system maximum drawdown (MDD) based on a backtest
When you run a back test, you’re back testing on historical data and like it or not there’s going to be some curve fitting in the system design. So when you run the back test and you determined that the maximum historical drawdown in your back test is 30% and you’re comfortable with the 30% drawdown, that’s all great provided the future is no worse than the past. But the trouble is what do we know about the magnitude of shocks in the market, about the potential for bigger and bigger uncertainties or panics?
The biggest market shock is always ahead of us!
Like there’s always a potentially bigger panic in the future. It may not come in the short term, but if you trade long enough then you are always going to exceed the maximum historical drawdown that your system achieves in backtesting.
When you run a back test on your system and you determined that the maximum historical drawdown was 30%, you want to be comfortable with a drawdown bigger than 30% in order to trade that system, because chances are something worse could happen in the future.
As a rule of thumb, your worst case could be the max historical drawdown x 1.5
A useful rule of thumb is to backtest your trading system and look at the maximum historical drawdown times 1.5 and ask yourself whether you could I tolerate and survive that. If you can’t survive that with your current capital allocation / leverage / position size level then you should adjust your portfolio management rules to ensure that you can.
This is a great way to think about how much capital should you allocate to a system or to a market direction or to a suite of systems. For example, if you have all loan side trend-following systems and they’ve all got a maximum of 30% historical drawdown, then you know at some point in the future, you’re probably going to get a bigger than 30% drawdown. Maybe therefore you shouldn’t allocate all your money to those systems.
Using Monte-Carlo Simulation to Better Understand Maximum Trading Drawdown (MDD)
How do you determine the maximum likely drawdown based on your back-tested trading system? Well, the first thing you’re going to do is you’re going to back test your system and see what you got in the back test. As a general rule of thumb, as stated above, if you multiply the maximum historical drawdown by 1.5, then you’ll get a indication of how bad it really could get.
Of course as a trader you are going to hope it’s not going to get that bad, but it might. So you want to be able to survive that level of drawdown from your system.
There are some other tools that you can use because the shape of your equity curve in the past, in your back test is determined by the sequence of events that happened in the past. But a different sequence of events could play out in the future.
The sequence of events in the future could be quite different… You could have this sequence:
Or you could have a completely different order of market conditions like:
These are going to have completely different drawdown profiles even though the overall outcome is exactly the same, so how can you determine how bad your drawdown could get?
One thing you can do is run a Monte Carlo simulation on your equity curve. The functionality exists in AmiBroker to take your equity curve and reshuffle it. This is an advanced topic, which I teach in my advanced back testing and trading system development course.
What monte-carlo simulation of the equity curve allows you to do is randomly resequence the daily changes in your backtest equity curve to see how good or bad your system could get when the daily changes happen in a different, randomised order. When you do this analysis you can see how your portfolio equity curve might look with thousands of randomly generated sequences of trading days:
Within each of these thousands of simulations, the maximum historical drawdown will be different. So by analysing all of these equity curves you can get a probabilistic understanding of the range of possible max drawdowns from your trading system. This is illustrated by the probability distribution diagram below:
The above monte-carlo output shows that you have a 100% chance of having a drawdown of more than 23%… a 50% chance of having a maximum drawdown of more than 30% and a 10% chance of having a max drawdown of more than 40%.
By doing the monte-carlo reshuffle of the equity curve you are asking how bad it could get if you re-sequence things. This is obviously a much deeper level of insight than just a single equity curve from one backtest.
Try and think, through different sequences of events, different types of markets, what would this system do? Do that logically, and then do it with a Monte Carlo simulation.
The relationship between drawdown depth and drawdown duration
The next aspect of drawdown I want to talk about is the relationship between drawdown depth and drawdown duration. There’s some schools of thought that say the depth of the drawdown can be reduced, but that’s usually at the expense of the duration. I don’t necessarily think this is a hard and fast rule, but there is definitely a trade off between the drawdown depth and duration.
For example, let’s say you turn off your system when the market starts dipping and you turn it back on when the market is rallying, you’re going to have a long, shallow drawdown. But if you don’t turn off your system when the market is dipping, you may have a big drawdown. But then when the market rallies, you’re going to recover more quickly. So the trade off is a short, deeper duration verses a longer, shallow drawdown.
This is a trade off that you can investigate further once you have learned how to backtest and optimize your trading systems using the skills that I teach in The Trader Success System. When you are backtesting different ideas to reduce your max drawdown you need to think about the impact on both depth and duration of drawdown and monitor the impact of each on your system performance.
This is the reason why often turning a system off and on again, based on what the market’s doing is sometimes not that helpful because you just end up with a really long, slightly shallower drawdown. In some cases you can get better compounding if you don’t turn the system off, you just trade through the drawdown and you get out of the other side to new equity highs more quickly.
Lower drawdown is not necessarily better if it’s twice as long!
So think about drawdown depth and duration and observe these when you make changes to your systems and that will help inform your decisions on how to improve the trading system.
How do you turn off a trading system to reduce drawdown?
You could introduce a new trading rule in your system to turn the system off if the market was in an unfavorable condition. This should be a systematic / objective rule and not a subjective judgement. A simple example could be to only buy if all all of your buy rules are in place AND index is above the 200 day moving average. Plus if the index drops below the 200 day moving average then you close all your positions. This basically means the system is ON if the Index is above the 200 day moving average and it is OFF if the index is below the moving average.
This sort of index switch / market regime filter sounds like a good idea, and sometimes it is, but sometimes it is not a good idea. You have to backtest it and evaluate the rule objectively by looking at the drawdown depth and the drawdown duration of the two scenarios and decide, which is best.
The causes of drawdown in stock trading systems:
There are several big factors that cause drawdowns in your stock trading systems. When you run a back test and you find that the maximum historical drawdown is too big, you need to figure out which of these causes is the problem so that you can go and address the correct cause to make the system better.
Experiencing a long string of losing trades
The first possible cause is a long string of losses. Many losing trades in a row caused by a entry signal that’s out of sync with the market or is no longer working. If you get many losing trades in a row, you can definitely go into drawdown and it can hurt your trading system profitability.
So sometimes improving the drawdown or reducing the drawdown is about making the entries trigger more accurate to reduce the length of the longest losing streak.
A broad market correction can push all stocks down and cause a drawdown
The second potential cause of drawdown, which is fairly common, is a broad market correction, which has all of your stocks against you. Let’s say you’re predominantly a long side trader and the market dip suddenly all goes into a sudden unexpected bear market (Hello coronavirus?!?!). Then that will push all of your positions down, which will cause a drawdown. This is probably one of the most common ones where you’re basically heavily invested on the long side. You’re therefore subject to the whims of the market.
So how do you deal with that?
Reducing this sort of drawdown is about identifying when a market dip is happening and trying to create a rule that will get you out or get you out of the weakest trades, or stopping your system from taking new trades. You don’t necessarily have to turn the system off, but if the market starts dipping then stocks which have a high beta fall faster, then you probably want to get out of those. However some stocks might hold up and actually not be worth selling because as soon as the market starts to recover they sort of take off again. This is something that you should backtest to find the best course of action objectively.
Inefficient exits causing a large difference between Maximum Favourable Excursion and Final Trade Profit
Third cause of drawdown, which is actually surprisingly common is a hugely profitable trade, which falls a long way from its peak to the entry price.
So this is where the comparison between the maximum favourable excursion of the stock and the ultimate profit of the trade becomes important. When you run your back test, you can sort on maximum favourable excursion and compare the MFE to the percent profit per trade and see if there’s a big disconnect. If there’s a big disconnect, can cause higher drawdowns than you’d like. This is a sign for you to look at the efficiency of your exit and see if you can design a more efficient exit to capture more of the MFE in these large moves. This is a great way to improve your profitability and reduce your system drawdown at the same time.
Entries which buy into weakness when the broad market is falling
The fourth common cause of large drawdowns which is typical for mean reversion systems, are entries which buy into weakness, that keep buying when the broad market is falling. For some mean reversion systems, that can be a serious problem. The more selective the entries of your mean reversion system, the less this is a problem.
If you have a very generic oversold entry for your mean reversion trading system that is easily triggered and generates lots of trades, chances are at some point when the market rolls over, you’re going to get fully invested in one day, then get stopped out, then get fully invested the next day, then get stopped out, then get fully invested the next day, then get stopped out. When this happens you will find yourself in a big drawdown hole, a big one.
The solution to this problem is to backtest more selective entries and ensure that your system does not keep entering immediately after being stopped out.
How to reduce your trading system’s maximum drawdowns
Reducing your trading system maximum drawdown comes down to an investigation and hypothesis generation exercise. You need to look at the backtest statistics and the equity curve to figure out which of these causes is really the problem with your trading system. Then come up with rules to mitigate it in the particular situation that your trading system is in.
In particular you can look at the following potential solutions to the drawdown problem:
- Improve your entry trigger to reduce the length of the longest losing streak
- Test a market filter for both entries and / or exits
- Look at the maximum favourable excursion profile of your backtest and see if you can make a more efficient exit that captures more of the profits
- Refine your mean reversion entries to ensure they do not buying into a market wide correction
- Ensure you are not buying and getting stopped out and buying again (eliminate this by enabling a reentrydelay in your trading system rules)
Investigate the causes of the large historical drawdowns and find the issue that is causing the drawdown that you’re uncomfortable with. Then generate hypotheses to address that and then test them one at a time and see if they improve the situation. That’s how you reduce drawdown in a structured way.
What you shouldn’t do is just throw indicators into your system and randomly backtest them to see if they reduce the drawdown!
Have a look at the system, have a look at the back test performance, have a look at the equity curve, figure out which issue is causing the drawdown and it will point you to which part of the system needs to be addressed.
If you want to learn how to objectively and systematically evaluate your trading rules, become a profitable and consistent systematic stock trader then click the button below to learn more about The Trader Success System – It will change the way you think about trading and transform your results!
Do you have any questions about how to manage and reduce the maximum drawdown from your trading system? Comment below and I will help you out 👇👇👇
- Learn 4 Causes of drawdown & 5 ways to reduce maximum drawdown in your trading
- Stock Trading Systems: What Are Reasonable Cagr And Drawdown Targets?
- How to Manage Initial Drawdown – Two methods to start a new stock trading system