Transcription
I back tested 100 trading strategies and I found this. Here are just some of the example exurs from the hundreds of back tests that I've run. Now, this has been on futures, crypto, equities, options, pretty much all markets. And over this period of time, I've wondered, and maybe you've wondered as well, what actually makes a trading strategy work on live markets.
There's a very big difference between something working on a back test compared to a live market. As you probably know if you've done any back testing yourself. Now, I used to assume it was all because of discipline and a sense of market experience that I just didn't have, especially when I was in my first year or first couple of months trading. I really thought all of profitability was just up to discipline really and men's experience.
But after back testing hundreds of trading strategies and consuming countless lectures, books, interviews, courses, and trading myself for the past 6 years, I've made some very surprising discoveries like how conventional advice is mostly wrong as well as two straightforward ways that I became profitable in trading. So what does actually hundreds of back testing trading strategies actually look like? Well, here's just one of the folders from my back testing data, which was about 45 GB of data. Now, in relative terms, it's not that huge, but at least in the sense where this is a lot of data files and also a lot of obviously code files for strategies. It took a lot of time to actually get all the data, the code, and also revisions on those particular strategies.
Much of what I'm about to share comes from those ahead of me. So, this is from lectures, books, and interviews. Just to give a couple shoutouts, these are three great books. The first book is around the story of Jim Simons, which is one of the most profitable quant hedge fund managers pretty much ever. The next one is a book by Robert Carver, which goes over futures trading strategies and how he approaches them. And then the last one is a bit more technical, uh, but I wanted to include it as well, which was testing and tuning market trading systems by Timothy Masters. Really great book, even if you don't code in C++. Uh, still actually a pretty good book to check out.
But this wasn't all just from books. This has also come from my own experience trading profitably. And this was my last month's equity curve where I run a portfolio of 10 automated trading strategies on the futures market. These are things like ESNQ, etc. So, it's actually been very easy for me to see what has worked and what hasn't without having to manually trade all these strategies by hand because a lot of them are just automated. Now, I would say I'm fairly profitable, but I'm not seven figures yet, but I do manage low six figures of capital of my own personal money. So, not through prop firms or anything like that. And probably over this time, back testing has been the biggest help as it showed me what has worked and what didn't.
So, let's get into the first discovery. If I just zoom out here real quickly, the first big discovery that I really uncovered over this time was that simple beats complex pretty much every single time. About 95% of the strategies that failed had one thing in common. They were overly complex and had way too many parameters. I'm talking about 10 maybe 20 parameters. Whereas, the strategies that actually made money, they had about three to five parameters maximum, and they had clear logical rules. The more you understand the reason your strategy makes money, the better. To give you an example, does your strategy exploit market inefficiencies? You need to understand whether you're capturing a risk premium, so getting paid for taking on systematic risk, or you may be exploiting something that's a pure inefficiency, let's say something like arbitrage, for example. And if you can't explain your strategy edge in a couple of sentences, it probably means you don't have one.
Now, that isn't in all cases. Some people definitely can have more complex edges. But if you're taking it as a large sample size, I would happily bet that most complex strategies are going to fail a lot more often compared to simple strategies, especially with the difference of that complex strategies will often look amazing on historical back tests, but they'll completely fail on live execution. Now, they will actually traditionally normally fail more advanced back testing approaches. Now, if this is your first time getting more into systematic or proper back testing, you may have not heard of them, such as things like Monte Carlo simulations or parameter sensitivity or out of sample testing. Now, those are normally very effective ways to test more of these complex strategies and get a sense of if they're going to fail already on live markets before actually deploying them. But ultimately, if you try hard enough, you will make a strategy look amazing on a historical back test if you just keep on optimizing. But it definitely will not work on live markets. And for this reason, most of my strategies are incredibly simple, as well as many other amazing quant and systematic traders I know. They all trade very simple things. They understand what they're trading and why ultimately they are getting a return. And that's probably one of the most important factors to it.
Now, the next discovery was a bit more of a long one, but if I just delete this box here, it was that your exit strategy matters 10x more than your entries. This was probably the most counterintuitive lesson from my back testing. I've tested many trend following systems with random entries versus optimized ones. Now the difference between those have normally been about 10 to 15% of those actual returns. Then I tested random exits versus systematic exits on the same entries. The difference was over 200% in rate of returns. Now this difference is higher in trend for strategies compared to others. But a similar trend does continue over many strategy types.
Now I'm just going to take a second here to dive a bit deeper into that as some of you may have questions. Why is that? Why do your exits actually normally matter more than your entries? The main reason being is in a lot of strategy types such as trend following, what you're actually truly capturing is the trying to capture the meat of the move, let's say. So if this was a general price trend, you entering here or here or here doesn't ultimately matter that much as long as you are exiting as much at the extreme or capturing most of the move as possible cuz that's where most of those returns are actually going to come from. Now, this is a bit different for mean reversion strategies because mean reversion strategies, you're more trying to focus on a higher win rate, but still then if you are nailing your exits there, you're having a higher consistent win rate and you're not having trades come back onto you. So, it's the same sort of trend that exits for the most part matter more than your entries. And I would recommend if any of you are already back testing your strategies, I would recommend trying to randomize your entries and see how much that actually affects performance compared to random exits and see if you actually have something that could be called some edge uh in your particular strategy. And this is why you can often extract way better returns with better exits than better entries.
And this isn't just relevant to me. As Jerry Parker said, one of the original title traders who has actually made I think hundred million plus dollars now explained the whole point is to take small losses and let profits run. Now that he is talking about trend following in this example that isn't the exact same with mean reversion but the point still stands where the exits are one of the most important factors compared to entries. But ironically most of the traditional advice for traders is to do the exact opposite. They try and optimize the entries obsessively using arbitrary exits really and more focusing on how to get the pin perfect entry point where price goes instantly into your favor. Now I would argue if you spend 70% of your time thinking about your exit rules and 30% on your entries your P&L will often thank you in the future.
Now, trading profitably isn't about finding the perfect setup or having complete nerves of steel and perfect discipline. It's about building strategies with clear logic and diversifying them into a portfolio. If you want to check out another video of mine, it's called about the holy grail in trading, and it goes about building a portfolio of systematic strategies and why that's often the best returns and risk adjusted returns. Now, most traders fail at this because they're gambling on gut feelings instead of capturing systematic profit. Check out the first link in the description for the exact process I go through to back test and automate my particular strategies.