The Most Profitable Trading Pattern You Will Ever Encounter

Trading Systems Free Forex Trading Systems
899

[B]Optimal f[/B]
The most important work on money management for trading done by Ralph Vince. It will be the most important work on trading you will read, if you understand the points he’s putting across. It is a very, very tough read.
I spend three months reading that book and I read every paragraph at least four times because I had a very difficult time understanding his points. Ultimately I did and I’m grateful I read it.

[B]What is optimal f[/B]
Well it is a modification of Kelly. We said that Kelly will always be expressed as a fraction, in our case 0.25789.
Optimal f stands for “optimal fraction”; the most accurate fraction of the kelly formula for betting.
The problem with kelly is that it provides the mathematically correct fraction we should use in trading, [B]assuming there is only two outcomes to the trade[/B]. If we can only win $100 and lose $50 for the example.
As we know this is not the case in trading; you will have different results. So optimal f addresses that.

I will not explain how to reach that, it is very complicated and difficult. I just want you to know that the optimal f is 0.4.

The next step would be to divide the worst loss by optimal f to give you the $f value. So, 1053/0.4= $2632.5.

Once you have that; Lot size= Balance/$f.

So I applied the strategy with the same starting balance of $10,000. It ended up with $43557.11. The highest it got was $55286 and the lowest was $5977.

It is important to note that the worst trade of $1053 was present in the trade logs. That means that at some point the optimal f did lose 40% of its equity on that trade and Kelly lost something of 25% on that trade. Yet sticking with that money management formula still resulted in 3X and 4X the starting balance.

[B]My take[/B]
I just take an average of Kelly an optimal f. the result is 0.328948997. The $f is 0.328948997. You can divide your starting balance by $f to get your lot size in mini pips.

The result was a profit of $39,221. The highest was $43,000 and the lowest was $6400.

That will probably be it for today. There is plenty more to talk about in the coming days. Until then, let me know if you have any questions.

900

Lots to digest. Thanks a lot for all of the information share!

901

Hey Doug when you get the time, I’d like to see a picture of the GBPCAD trade long the EA entered on January 20 and is still open. It is the only open trade.

The reason I’m asking is because I entered that trade earlier this year, in fact I had an explanation of its chart in my very first posts. I exited that trade a while back (my most profitable trade this year.) Yet for some reason the EA is still in it. I exited in late February.

902

Philip, I’ll be happy to show it to you, but probably a waste of time, since the 11-year data set ends on 1/21/2015 at 12:00 - in other words, that’s the last 4-hour bar I have, only a day after the trade fired. That’s why the Trade is still OPEN. Do you still want to see the chart of the setup?

903

No its ok. Thanks for clarifying that.

904

Philip, reading over your money management posts, which were excellent BTW, I do have a question. It appears to the naked eye, that these rules, some of which I wasn’t familiar with (Kelly, optimal f) derive the optimal Lot Size based on account balance, worst trade, and resultant calculated “factor” using back-testing sample size. Hence, taking trades with a specified Lot Size will result in different $ allocations across all CCY Pairs that are traded, naturally.

So, is Kelly and Optimal f calculated on a per CCY pair basis, or is it just the way it is? Am I missing something here?

905

I wonder if not all ccy pair failed trades are equal. for example it might be that a particular ccy pair performs badly more often than all the others. so maybe one should take the worst trade from every ccy pair and average it somehow… was never good at math, lol hope what i said makes sense and is along dough’s train of thought

906

This is kind of important. It is also similar to the point molonist put across. So let’s take it step by step. We need to answer a number of questions. What am I going to trade? What are the backtest results? What is my fraction?

So you need to know beforehand which instruments will you trade. If the answer is gold for example. You do the back testing on gold only then find its fraction.

If the answer however (and this is to address molonist’s point) is Forex, you have to do the back testing on all pairs and not a pair-by-pair method. Why? The most important reason is you want to know the absolute worst result that happened on any forex pair to use it in your calculations. But if I see that the worst in EURUSD was a 700 pip loss and trade accordingly, it can happen that I encounter a 1000-pip losing trade in the future. But when I do it on all 27 pairs for 11 years, it is a much more representative sample.

Doug we define our lot size based on $f value and not the other way around. You can change the balance but you can neither change $f nor the lot size.

Kelly and optimal f answer a very specific question? what lot size should I assign a trade based on the past performances of the trading system?

So if we change that lot size then we are not using kelly or optimal f. Optimal f in particular argues that it is THE RIGHT lot size. So if you change its lot size you will never make a higher return.

Its about to get much more confusing so bear with me.

907

I finished the results for the system with 27 pairs over the past 11 years. The winning % remained at the same areas, 59% counting break evens as wins. Win-loss ratio is at 1.14. Here are the results

Kelly Criterion: Ending balance was $15,305.

Williams’ formula: $15348.

Optimal f: $15501.

My average: $15432.

All of them never went below $7500
So how on earth, even though it is the same system, almost the same winning%, slightly lower w/l ratio but the result are much lower than the earlier results? Well for two reasons. I will answer it in my next post since I was going to address it in my money management post anyways.

908

@ Philippirrip - thanks again for your awesome system.

Can you recommend some good books/online resources for : technical analysis, forex trading education, research.

Further, I would like to ask you if you are a professional trader?

Also, can you suggest some good system for day-trading or even scalping?

909

This was meant to be the last part of the series but I moved it up since the questions were related to it. So let’s get going.

[B]Risk of ruin[/B]
Is a mathematical concept used in gambling, finance and trading as well. The risk of ruin addresses the odds that the next trade will result in incurring a loss so great that continuing trading afterwards will never be profitable.

In other words, it measures the probability of reaching 0 or at the very least a draw down level that could never be recovered in the future.

Here is the kicker: mathematicians discovered that the risk of ruin has a positive correlation with the number of trades; [B]The more and longer you trade, the more likely you will crash your account. [/B] This is a mathematical truth, keep trading and one day you will crash your account. In fact the more you trade, the more inevitable crashing your account becomes.

In his book [I]Portfolio Mathematics,[/I] Ralph Vince concludes in page 412 “To confirm the reader’s burgeoning uneasiness at this point consider the following: In the long-run sense, the probability of hitting a drawdown of approaches certainty as you continue to trade.”

So basically the fractions of optimal f and kelly decreased when we increased the number of bets because of this phenomenon (also, it encountered a bigger worst loser). If we trade even more, the fractions will get smaller and smaller.

The good thing about that is even after 11 years and 850 trade the money management accounts (all four of them) never faced a ruining drawdown, they also continued to make a profit. At some point, after 2000 or 3000 trades for example, we will have to crash your account.

So the measure of a good trading system in the longterm is not how much it makes, but how long it survives.

[B]Warren Buffet has been investing for 60 years yet he makes more money with time, doesn’t that proof risk of ruin wrong?[/B]

The answer is no. Warren Buffet is an investor, we are speculators. An investor basically becomes a partner in a business. What Buffet has been doing is finding companies that will be profitable and he buys a steak in them. So with more time, those companies make more profit and Buffet gets richer.

It is the opposite with us, we are speculators. The more we speculate, the more inevitable it will be that we crash our accounts.

But don’t be sad, I have a few ways to limit this truth. They are the subject of my final post on money management.

910

Unfortunately I cannot recommend you good books on trading. To be honest I read a lot of books but the knowledge I have was basically the result of learning from the mistakes of these books.

However two books that could give you a good system to follow are those two: 1) Larry Williams, Trade stocks and commodities with the insiders. 2) Robert Miner: High probability trading strategies.

I do more than trade forex :slight_smile: If I was a professional forex trader I would be trying to charge everyone for this system. I would have made a lot of money by the way too :P. But this is not what I’m about.

I would say 70% of my income comes from the financial markets. Basically forex, stock investments, bond investments and a small business is what I do.

The good system I recommend for scalping and day-trading is the “don’t do it” system. Never lost a dime with it.

911

So what do we do to limit risk of ruin? Two things: 1) Be selective in your trades. 2) Re-invest in your account.

  1. Be selective in your trades.

To trade using aggressive strategies you can only open one trade at a time. So set yourself a target number of trades and consider your “forex year.”

For example, I plan to open a new account to trade the optimal f and will post on y myfxbook so you guys can see it. I will set my self 24 trades at an average of two trades a month. So I want to take the best trades possible. These best trades for me are usually based on fundamentals. Or just have an RSI divergence filter like JW is doing for example.

Once I finish my 24 trades, it doesn’t mean I don’t trade for the rest of the year. But I have to rebalance my account. I don’t withdraw any money before that period. But I must withdraw some money after that period.

  1. Re-invest in your account:
    After those 24 trades, I re-invest my optimal f into the account and withdraw the rest. For example I applied the trades I took this year to this strategy, by the 24th trade I had a balance of $23475.2 (after deducting spreads and commissions and swaps.) The original balance is $10,000.
    Now I must rebalance the account. Since my optimal f is 0.22. I multiply my final balance by my optimal f. so 23475.2*0.22= 2964.544.
    So I will add $2964.544 to my original balance of $10,000. The remainder $10510.65 is withdrawn as profit. This already means I cannot lose my original investment. Then hopefully by the time my risk of ruin becomes inevitable, I would have made a huge return on my original investment.
912

@PhilipPirrip - great sound money management strategy! I love what I have read so far. With that said, drum roll… I have some cool exciting features I’ll be adding to the EA before I released the finalized version. This will include : autodetection of broker spread, autodetection of broker digits, email statistics options including a drop down to be able to pick when you’d like statistics sent, and… more I am looking into. I will update everyone on what I find out.

913

Does anyone know if RSI added to trading is working? Anyone got positive results from it?

914

What Buffet has been doing is finding companies that will be profitable and he buys a steak in them. So with more time, those companies make more profit and Buffet gets richer.

It is the opposite with us, we are speculators. The more we speculate, the more inevitable it will be that we crash our accounts.

Hmm… I’m not sure that I understand the distinction you want to make between investing and speculating. I’ve been dabbling in probability and inductive logic for awhile now. If you want to say that in the long run (infinite time and infinite trades) one will eventually crash his account because the probability of reaching a point of non-recovery is greater than 0, I agree with you. As Keyes said, “In the long run we’re all dead.” But Buffet cannot escape this either or at least I do not see how he could escape this.

Just as Buffet is able to find companies that WILL BE profitable, we also are attempting to find trades that WILL BE profitable. that ‘will be’ is not a certainty its a probability right? both for buffet and me. he’s in the same boat as we all are as far as I can tell.

915

The results we have are based on the monthly RSI filter

916

Techy, the results that I posted were less good when RSI Monthly filtering was removed. Also, when Swing Hi/Swing Lo were used as Initial Stops, the results were less good than when we had these Initial Stops disabled. My last posting of results had:

  1. RSI Monthly Filtering set at 66.67/33.33
  2. Initial Stops disabled

I also ran 1 simulation with RSI Monthly at 70/30, and there was no real appreciable difference, when I ran it against 11 years of all 27 CCY pairs.

917

Well we are agreeing. What I’m precisely saying is as we increase the numbers of trades, the odds of crashing our account reaches 100%. There is a formula for it that I’m happy to share. But it is such a complicated topic that I’d rather not discuss lol.

Ok let me try to rephrase on buffet. Warren is buying a cow. That cow produces milk. It can give birth to other cows that will produce milk and he could even use them after a while to produce meat and so on. It doesn’t matter to buffet whether the price of the cow is $1 or a $1000. He only cares that the cow keeps on producing milk.

On the other hand we don’t own any cows. We are just betting that the price of a cow, based on the milk its producing, will increase from $1 to $500. If that doesn’t happen we lose. Buffet wins if the price of cows goes down because he will buy another cow, he only cares about the milk. He doesn’t care about the price of cows.

Additionally, Buffet employs something calls the margin of safety. He buys a company below its actual price. So if the company goes out of business he still makes money. No body knows what method but the main idea is to buy the stock for a price below its book value. This is something we cannot do as speculators.

I hope that makes sense.

918

Phil… quite some interesting stuff you posted here. Barely came across any online literature going this deep (or maybe I didn’t check too much).

The optimalf got my attention. But i don’t quite see how it fits.

The next step would be to divide the worst loss by optimal f to give you the $f value. So, 1053/0.4= $2632.5.
Once you have that; Lot size= Balance/$f.
So I applied the strategy with the same starting balance of $10,000. It ended up with $43557.11. The highest it got was $55286 and the lowest was $5977.

From the above calculation, that would result in 3.79 lots per trade? Isn’t there a risk of risking too much?

Also, let’s just assume a new trader now, went in with 3 lots on the recent AUDUSD loss.

Assuming long position was entered at 0.78362, the trade hit a SL at the lowest low before the EMA crossover at 0.75594. That’s about a 278.6pip loss.

At 1 lot, 1pip equals $10, at 3 lots, its $30. So a 278.6pip loss = $8358 @ trading 3 lots.

Could you elaborate on this?