Over 5 MILLION Tests PROVE This Trading Strategy Works!

I've just spent the last two months running thousands of backtests on the most popular trading strategies and I think I can safely say I found the Holy Grail, the number one top performing strategy that gives the best results with any entry in any market. I'm going to give you the rules for you to copy, but more importantly, I'm going to explain exactly what the edge is that separates this from all the others. So you can really trust it. We've got to start by looking at what is a trading edge and why most trading strategies fail. There is an uncomfortable truth that retail traders often struggle with, and that is that all financial markets are mostly random. They're not entirely random. There are inefficiencies and non random repeating patterns. But the vast majority of all price moves close to 99% are random. And it is mathematically impossible to profit in the long term from random data. If you have a simple 1:1 profit to loss ratio, 50% of the trades will gain 1R and 50% will lose 1R. This is identical to a random coin flip where you win $10 for heads and lose $10 for tails. The wins and losses are equal in size and quantity and they cancel each other out for a return of net zero. With trading, we can alter that ratio. Instead of a one to one ratio, we can change it to one to two, for example. But in random market random data, any adjustment to the ratio proportionally adjusts the win rate. We will win two out of three trades, but the losses are twice as big and the net result is still zero. It makes no difference how much you adjust those ratios or in which direction, or all results cancel each other out. And it's impossible to profit. Scaling out and taking partial profits won't alter the outcome. Neither will scaling in and adding two positions. Each individual sub portion of a trade is constrained by the same mathematics. So no matter how you combine them, when the individual results are 0 combined, they are still zero. Even using a fixed distance trailing stop, loss makes no difference. Moving the stop simply adjusts the ratios and the win rates. Like all the other examples, it is a mathematical impossibility to make positive returns in the long term from random markets. But it's worse than that in trading because of costs, spreads and commissions. Those costs tilt the win loss balance slightly more to the losing side, meaning we will lose more often and the losses are bigger than the wins. So what was a net zero result now changes to a net zero negative. In random markets with trading costs, all strategies are losers, so you'd better Be damn sure you have an edge that exploits the non random price moves and patterns in the market. Without a real edge, if you are trading the random elements of the market, you are mathematically certain to fail. And most systems and traders do fail. So how do you test your system's edge? Well that's where backtesting comes in, right? Wrong backtesting. Historical charge in your broker platform is meaningless. In fact it's worse than that. The results are always false. It either won't show you the edge on a profitable system or it will produce a fake edge in a losing system. You will never get an accurate result doing normal backtesting. Let me explain. This is a six month backtest that returned 43% profit. And this is another six month backtest that lost 26%. One has an age, the other one doesn't, right? Well they are the same system on the same pay over the same time period. So what's going on? The first test entered at 8:00am and the second entered at 8:01. So does that mean entering at 8:00am is better than 8:01? Do you lose your edge if you're one minute late to the entry? Well here's the same system again on the same pay over the same time period. But this time the entry was 8 16, this one entered at 1230, this one at 1337 and this one at 1500. And just to mix it up, the same system from the year before. And this is from 5 years ago, both with an 8am entry. Does this system have an edge or not? Well this is the flaw with backtesting. If the entry is here, this pullback never quite reaches the stop loss and then the price continues up to the profit target. A slightly later entry, that pullback would stop the trade out for a loss. In live markets. You and I could hit the buy button at exactly the same time, but depth filled at a slightly different price due to a split second change to the spread liquidity levels, depth of market, our different Internet speeds or time latency caused by the distances between our locations and our broker servers or our broker service from the market service. And even if the difference in our entry price is only a tenth of a pipeline, that can be the difference between one of us getting stopped date and the other reaching a target. We will get entirely different results from the same system because of the conditions being slightly different fractions of a second apart. And those differences are random, chaotic, always fluctuating and never the same twice. So I'll ask you again, does this system have an edge What? Which one of these tests is more accurate and trustworthy? Which one is representative of the results that we can expect in the future? Can we expect positive results in the future? You probably have no idea. I certainly wouldn't. But there is a way to test this properly and determine the real edge. And to show you how, I need to quickly demonstrate using examples from a random number generator. This spreadsheet simulates a series of 200 trades with the wins and losses randomly distributed. That's roughly equivalent to about a year's worth of day trading. And I've used the settings. With a 49% win rate and a 1 to 1 reward to risk ratio, this is a guaranteed losing system. And yet, each time I refresh that page to generate a new sequence, sometimes the results are positive. Some of them are fantastic. Now, a moment ago I said it was impossible to profit in random data. And yet there are multiple profitable results here after 200 trades. What's going on? Random distribution. It is impossible to profit in the long term. But in the short term, thanks to the random distribution of wins and losses, the random sequence that they come in, and the random clusters of consecutive wins or consecutive losses, sometimes you can get positive results. In fact, quite often. Now think back to what I said earlier about the markets being mostly random. If you test a small sample size of history, such as six months, you are getting random results. Which is why that system that I tested came back with wildly different returns. Each test was looking at a small one off sample of randomly distributed wins and losses, exactly like these spreadsheet trials. So you might be thinking that the solution is to test a larger sample size. Wrong again. A larger sample size will certainly get us closer to the real edge of the system, but it still isn't accurate and can't be trusted. I'd now change the spreadsheet to generate 1,000 trades, which is roughly equivalent to five years of day trading. And this time I'd got to refresh that data more times to occasionally produce profits. Most trials are losses, as we would expect with a losing system, but even so, there are still multiple trials that generate positive returns. Random distribution is something amateur traders just don't understand. It's easy to think that a guaranteed losing system would be guaranteed to lose after a thousand trades, but there are plenty of sequences here that end in profit. I can even increase the number of trials to 4,000, which is closer to 20 years of trading data. And while the results are much closer to the true edge, with this larger sample size, and almost every time I refresh that page. The results are loss after loss after loss. If I keep doing this enough times, there are the occasional profitable series and even some very impressive profits. Imagine if this was the only test I did, thinking that 4,000 trades is more than enough to find the edge. Of course I'm going to assume this system has an edge when it doesn't. And this is the real problem with historical charts in your broker platform. They are a one time set of data. Those exact conditions will never ever happen again. When you backtest the system on that one frozen set of numbers, you have no idea if the result is representative of the real edge. One test is too small to determine any conclusions. When you begin to forward trade in the future live market, it's the same as generating a new random sequence of data, a new set of conditions, and the results you get will be different, perhaps polar opposite, from the last sequence generated. Just like hitting refresh in the spreadsheet. This is the reason for the legal disclaimers that state past results are not indicative of future results. The results you get from a historical chart are a one off and will never ever be repeated. Retail traders ignore that warning, thinking it's just some nonsense government regulation to be paid lip service to. It's actually one of the most important warnings any trader could ever hear. And this goes both ways, by the way. I'll switch the inputs to give this system a 55% win rate. So a guaranteed profitable system with a decent edge that will make you extremely rich over the long term. But with a sample size of 200 trades, it shows up as a loss almost as often as wins. If you only ran one small backtest and this was the result, you would mistakenly throw out a fantastic system simply because you didn't understand random distribution. Even looking at a sample size of a thousand trades. What if you started trading this winning system in live markets here and for the next six months or even a year you went into a losing period? You would naturally think something was wrong, the system had stopped working, or perhaps you were suddenly trading poorly and the likelihood is you would change something. You would give up on a winning system because you didn't understand random distribution. This is the Achilles heel of most traders. They don't understand random distribution. And the back testing they do on historical charts tells them absolutely nothing about any potential edge or expectancy of this system. So what's the solution? Well, sticking with the spreadsheet just for a moment. If I didn't know what inputs I'd used, the only way to find out if this system had an edge or not is to run multiple trials, hit refresh a thousand times and take an aggregate. If 60 or 70% of the trials produce profits, I know I have a real edge. If 60 or 70% of the trials ended in losses, I would know there is no edge. We need to take the same approach with market data and there are several ways you can do that. You could simply test the strategy on different instruments. So as a Forex trader, I can run the test on 20 different currency pairs. Some will return profit, some will return a loss. If it's profitable on more pairs than it's not, I can be confident I've got an edge now, particularly in Forex, but I'm sure it's the same in other markets. If you go to any forums, you will see people post a system and state it only works on the pound US dollar, for example, on the one hour chart. So don't trade it on any other pairs or time frames. If that is true, then they've got a losing system and it shouldn't be traded anywhere. What they've done is run the system across multiple series of tests and conditions like hitting refresh on the spreadsheet. And most have been losses. But they focused on the one and only run that came out positive and falsely concluded it works on that pair or timeframe, but not the others. That's an over optimized, curve fitted strategy that worked on one unique historical set of data and will never work again in the future on new data. To know if you've truly got an edge, you need to test the system on as much data as possible. A large sample size with variable conditions. Test different instruments, different time frames, different slippage and different spreads. Ignore the individual results, but take the cumulative total. If 60, 70, 80% of your trials return positive results, you have an edge that will continue to work in the future under new market conditions. And you'll be pleased to know I've done all of this testing for you, which is why I can confidently say I've identified the strategies that don't work plus one that can, I think, legitimately be called a holy grail. I'm going to show you all of these results now. One quick caveat before I go into this. I am not testing the entry in terms of trying to predict the next price direction. I know new traders are obsessed with the entry and predicting direction and they think that's where the edge is found. But I don't care about predicting direction, I only care about exploiting direction. We don't make money when we enter a trade, we make it when we close it. So what I've tested is, is how to manage and exit the trade once we're in it, regardless of what the entry is. Which is why I'm saying this can be applied to any system in whatever entry you prefer to use. So these were my tests. I entered a buy every 15 minutes, so buy at 8am, 8:15, 8:30 and so on throughout the full 24 hours. And what that does is recreate a different sequence of outcomes like I showed you earlier, with the difference of entering slightly different times. It changes whether and when a target level is hit or not, which in turn changes the random distribution of the wins and the losses. That gives me 96 different trial series to test. And I did the same on the sell side, so about 192 trials. And I test BIS on 28 different currency pairs, taking the total number of trials to over 5000. And the historical look back was 20 years per chart. That's the same as hitting refresh on that spreadsheet 5,000 times per strategy. So it is a pretty large sample size of tests. Then the strategies themselves were different reward to risk ratios like taking profit at 1 to 1 or 10 to 1 and different pip ranges starting with 10 pips up to 100 pips. So one strategy was enter with a 10 pip stop loss and a take profit at a 5 to 1 ratio. So 50 pips. Another strategy was to enter with a 100 pip stop loss and take profit at a 10 to 1 redshift. So 1000 pips and so on, including trailing the stock loss at fixed distances. There were a thousand different unique strategy combinations. Each one tested on over 5,000 different samples of 20 year tick data. That's well over 5 million tests in total. Not something you can do in a standard MT4 broker platform. You need professional data and professional software. And you can appreciate why it's taking me two months to complete these tests. I even had to rent multiple servers because my personal computer obviously wasn't powerful enough. Now, just like the tests in the spreadsheet, some of the results were positive and some were negative. Remember random distribution. But when they were aggregated, every single strategy was negative expectancy. In the long run, more than 50% of the trials of each strategy ended negative, so none of them had an edge. But there were patterns. Which will help explain the one final strategy that does have an edge. First of all, not all of those strategies were equally bad. Some came very close. The very worst strategies were the ones that used small PIP ranges and small ratios, entering with a 10 pip stop loss and a one to one reward to risk ratio was the absolute worst. And gradually the results got better as you used wider PIP ranges and bigger profit ratios. Using a 100 pip stop loss and a 10 to 1 profit target was the best. That makes perfect sense when you think about the random nature of the markets. Markets are mostly random and as I've explained, it is impossible to profit from random data. Let's say for the sake of illustration that 23 out of 24 hours are random and then for one hour a day there is a non random move. For you to profit in the long term you have to catch that non random move and keep catching it again and again across multiple days. The trouble is exactly what hour of the day that non random move happens changes from day to day. If you're an intraday trader whose trades last from minutes to a few hours, in fact anything less than 24 hours, statistically speaking, you are very unlikely to be precise enough or lucky enough to be in the market during the small window of time the non random move is happening. You are almost certainly only ever trading random noise. That's why the strategies that were in and out of the market very quickly performed the very worst. They were engaged almost entirely in random data that is impossible to profit from. And the ones that stayed in the market longer, either by having bigger profit targets or using wider ranges, performed better. The longer any trade remains open, the more chance of capturing the non random moves. If you leave a position open for six months and there's one hour a day of non random moves, you've accumulated about 130 hours of profitable price action. And that can make you millions. That's the fundamental reason banks hold positions long term and don't waste their time scalping. But all of these strategies ultimately fail to get across the threshold to have a statistically relevant edge because they all closed out too early because they use profit targets. If you enter a trade with a fixed take profit target, that's an arbitrary value you've decided in advance, which has got nothing to do with what the market does in the future. Or why take profit at 3r if the market is going to go 30r now when I talk about leaving trades open longer term, that does not mean simply trading off higher time frame charts. I could trade off a daily chart using a 3 to 1 reward to risk ratio and keep the position open for six months to capture the non random moves and at the end of the day have a positive expectancy and make money. But that's hardly worth it. If my stock loss on a daily chart was 500 pips, risking 1% of my account, my profit is 1500 pips, which took several months to capture. But it only returns 3% on my account. It's a waste of time. What I do instead is enter on the lower time frames like a 1 minute chart and scale up. And this is the final strategy that I tested and the only one that returned a positive edge. This is how it works. I enter a position with a very small stop loss of perhaps 10 pips, which still represents 1% risk to my account if the price goes in my direction. Instead of closing with an equally small profit target or trading with a constantly fixed 10 pip trailing stop loss, I adjust the stop loss and scale up the time frames. So let's say that initial 10 pip stop loss was based on the average true range on a one minute chart. Once the price has gone a one to one ratio, for example, I'll move the stop up to break even and then change the chart time frame to five minutes. The ATR on a five minute chart may be 25 pips. So I wait to trail that stop 25 pips behind the price. As soon as it improves on my break even level, I change the time frame to 15 minutes, where the ATR may now be 60 pips. So I change the trailing distance to 60 pips behind the price. Every time the stop loss is able to move up and improve on the previous time frame level, I scale up to the next higher time frame and gradually that stop loss is getting wider and wider and wider. Eventually, if I get up to the daily chart, my stop loss may be 500 pips behind the price. And I can keep this open for months without any concern about being stopped out from the random noise and chop in the short term. But unlike if I entered on the daily chart to start with VAT, 500 pips now represents 50% return instead of only 1%. It's not unheard of to double your account in a single trade like this. At the time of recording this video, I entered a position two weeks ago with an 8 pip stop loss. The current price is now 750 pips higher, giving me an unrealized profit of 93% on my account. If I closed out now, my trading stop is locked in just over 300 pips, which is still nearly 40% return. That's much more worthwhile, and you don't need many trades like that each year to make an awful lot of money. This is the only system that I've tested that demonstrated a consistent and robust statistically relevant edge. Now, there are some important considerations with this that retail traders are not always comfortable with. Most of the trades make no money. The entry uses very tight stops, which means it is in amongst all of that random short term noise and chop, just like every other intraday trader. So at best it's like a conflict and 50% of the positions are going to lose. Of the 50% that make it to break even, most of those will be stopped out for little or no profit. What that means is most of the time your account will be going sideways with lots of small losses and small wins canceling each other out. But to be fair, that's what most luteal traders are doing anyway. But what separates this strategy from the others is occasionally we'll hit the big trend where the magic happens. Believe me, it's easier to make a thousand pips profit in one go than it is to try and make 10 pips every day for 100 days consistently, which is what amateur traders typically attempt. As I said, I tested this on over 5000 trial series. Basically a random entry spaced every 15 minutes throughout the day. On 28 pairs with 20 years of chart history, 76% of the trials returned positive results. That's a rock solid, robust edge that will continue to work in the future in new market conditions. Now of course, I didn't just stumble across this blindly by accident. This was the strategy that I was taught when I joined the bank trading floor 26 years ago. The idea is to make lots of small, low cost attempts to get into the market, accepting that most will fail. But then when you do occasionally catch hold of one, ride it for all it's worth. And the edge in this is easy to explain. It works because the strategy leaves the positions open long enough to accumulate all of the occasional non random moves in the market or whenever they happen. Can you describe your edge so concisely as that? The important point here is not that this exact strategy is perfect or the only option. It's proving the concept. Keeping trades open over the long term is the key principle. How you manage that is flexible. You can use the ATR value as I've demonstrated, or you can use a moving average on progressively higher time frames or values. Or you may do it as a percentage of the unrealized profit, or if you're experienced enough, using your own discretion, perhaps looking at swing points. I personally use combination of the ATR and my own discretion. That's the hallmark of a robust strategy. It's not clear fitted or over optimized to only work with one set of data with one very strict set of rules. It works even when you change some variables because the overarching principle is is a sound concept, a solid foundation and remember this is a trade management strategy. You can continue using any entry you prefer and simply add this is the method by which you keep that position open and eventually close it. But one last word of warning no strategy or system will work for you if you can't trade. The system is just a tool you need to be skilled enough to to use it. You're the deciding factor. Trade in is not reliant on the system it's reliant on you and your ability to manage your decision making. So if you really want to improve your results take a look at my psychology course series. Work on yourself first then at that risk management strategy to suit your requirements with your system.