Is it better to increase lot size after win or after loss?

Hi,
suppose you have an ea with win/lose ratio above 50%
and high expectency
is it Better 2 increase lot size after win or after loss?
i do not mean double lot size like martingale
it is very risky,
i mean increasing it by 20% or so
please vote above ^^^^^

i think it would be better decreasing lot size after a loss as you have suffered a loss and your account will be lower in funds therefore you will need to adjust the lot size to get back your risk ratio

as for win it depends on how you have just won as to wether you can go up in size or not

regards damien

if the above is true then you should leave them constant (and slowly adjust to the account size every few months). There would be no need for martingale or antimartingale if you have a profitable strategy.

I created a coin toss indicator that has a margin of error of less than 1/100,000 of 1 percent randomness in 100 million coin tosses. (See screen shot)

By taking the Balance and dividing it in half and only using half of the balance to risk, and then setting the bet size as 100 percent of the half to risk, and by placing a full buy bet of 1 bet if the previous coin toss was heads, or placing a 1/2 buy bet of 0.5 if the previous coin toss was tails AND placing a limit of >= 200 for a max profit or a max limit of >= 25 for a max loss and having an starting balance of 100, the program runs.

Each cycle of the coin toss will go for a max of 100 coin flips, however, if the either limit is hit, the 100 toss sessions stops and a new one starts.

I performed 40 sets of 100,000 ( 4 million total ) of these sessions and the stats are listed in the log attached.

Note that:
AvgBal is the average balance at the end of each session of 100 tosses for 1000 of theses session ( a possible 100,000 tosses )

Accum is the accumulated balance of the sessions for each 1000 iterations of the sessions.

RunTotal is the running total of the balance for all sessions and tosses.

Avg I is the average number of tosses that occurred in each set of 100 toss sessions before either limit was hit.

Some interesting findings:

1. The account balance grew over the course of the test.

2. The draw down at it's most was -4350

3. The final Running Total was 23,313.45 (from a starting balance of 100 )

4. The average number of coin tosses in each set of a maximum of 100 coin tosses stayed around only 1 to 3 tosses before a limit was hit.

5. It was impossible for me to lose all of my balance, because I always divided it in half and then recombined it for the next toss, thereby always keeping half in reserve and never risking the entire account. This means that my risk of ruin is zero.

6. This was a completely random walk coin toss with a margin of error of less than 1/100,000 of 1 percent pure random.

7. I conclude that the reason that this is profitable even for a random coin toss is because this is taking advantage of very short term trends in random data. It is betting more when there is multiple heads and less when there is not and is stopping each of the sessions at a 8:1 reward to risk ratio. This is much the same idea behind Dr. Thorp's Black Jack method of betting and odds counting from the early 1960s, however my method of account control is very different, as he used the Kelly criteria for his approach. I believe if I had done so also with this test the results would be even greater.

8. This would suggest that there are periods of non randomness in larger areas of random data and would also suggest that increasing your bet to take advantage of this phenomenon is a winning strategy.

EDIT I made a typo above. I said 1 percent of half of the account to risk and I meant to say 100 % of half of the account to risk. I have corrected it in red. If it had been 1 % the results would be 100 time less, although, still profitable. Sorry for my typing error. The test reflects 100% of half of the account to risk. All the conclusions are still valid and not effected by this.

All due respect to those with education in probabilities, and statistics. There is one fatal flaw in applying gambling methods to markets.

All "games of chance" have a finite structure, meaning the probabilities are fixed and will not change. Thus, martingale and other similar methods seem to work better. However, market structure is NOT finite. The probabilities and odds change with every new trader and participant world wide. Add to that the possibility that other traders will enter or exit within any given moment, and you have an INfinite number of possibilities and therefore the odds are also INfinate.

The same methods that work in gambling will not work in markets because the method is built on the assumption that you will only lose x number of times before winning again. This is the fatal flaw. The market can make your system or method lose and undefined or infinite number of times causing ruination.

That is why I alway recommend a fixed percent of account on each trade with a STOP TRADING at a predetermined draw down point (if reached) every month or every quarter.

Thanks for your response here and I agree with you regarding gambling and trading, however that was not my point. This test stops precisely as you describe when a predefined loss or Profit is reached. I only use the random data to point out that the methodology is sound, not to make an argument for a gambling scenario. That was not the take away point that I wanted to present here. I am performing a type of market analysis. on a simulated market. You place a trade and it is either a win or a lose or a draw, those are your only options. So I have accounted for those options here.

Always risk a fixed percentage of capital on each trade. So increase your lot size after a win and decrease your lot size after a loss -> risking a fixed % each trade

I absolutely agree with this, although it is not optimum position sizing, it will always keep you betting with the winners and less when losing.

Your point is well taken and I too agree with your conclusions. MY only point is that many others will read these types of comparisons and draw parallels between your very well though out study, gambling, and markets. I only wish to point out that risk management is similar but not the same when trading markets. While the principle of uncertainty is common, the basic structure of the "game" is very different and can have disastrous results if not accounted for.

We all want to improve our skills and each of us contributes in his own way. As well each of us takes away something just a little bit different from these discussions.

No offense intended. I think your study shows very well what we are all aiming to achieve. Control of loss...maximization of profit.

LOL no offense taken. Who wants to talk with only those that agree with them? Nothing gained there. You may have noticed that while my percent of trade is high for presentation purposes of this test, at half the account and certainly not what should be traded in the real world, it is still a fixed percent of equity. So I am in agreement with you and Max on this point.

Thanks for the stimulating conversation, Scott

There are plenty of EAs with a high win %, say 95 and more but negative expectancy.

For these types of systems is at all possible to give them a positive expectancy by decreasing the lot size after a win?

Start of risking a high %, decrease after wins. When the loss comes it will be a small % of the account. Reapeat

Here's how I do it. Others may agree or disagree I don't care. I use the same lot size until my account is doubled. ONLY once my account has doubled I increase my lot size.

Then how does RunTotal become negative (at least twice)?
Please clarify.

IMHO the drawdown is actually much larger. You're carrying over your balance from one set to the next - hence

So your drawdown is at least the largest difference between highest RunTotal reached and lowest subsequent RunTotal, which seems to be higher than 7000 in at least two instances:
6545.77-(-529.23) = 7075
10331.51 - 2901.63 = 7429.88

(it is probably actually larger, because the extremes were probably reached during the cycle, not necessarily at its end).
More significantly, I think, is the percentage of drawdown - which exceeds 100%.

To me, this means that you could have ended up with a negative balance, too, and the fact that you ended up positive is pure chance, with no statistical significance.

I don't think that with a fair coin you could devise any net-positive system, regardless of the money-management system you choose to use.

You could demonstrate this by running this setup a few more times. Do you always get a positive balance at the end? I'd be surprised.

Please let me know if I've missed anything.
Thanks!

Just to add my 2 cents opinions

My actual code return triple digit gain over the last 13 months with less than 25% win ratio. When i win, I win a lot lot more than what i lost.

How do I manage my risk$;

Account grow by 5%, I increased my Risk$ by 5%
Account decrease by 5%, i decrease my Risk$ by 10% ***
*** Down to a minimum value. It wont go lower

NB: This is only use because my win ratio is very low

Let say that when you start you decided to bet 1% of your initial account. It is normal, as your account grow, to increase your risk$ to maintain the same ratio.

Using a system with over 50% win ratio i would

1- increase my risk$ by 10% when account grow by 10%
2- If account decrease by 35% i would decrease my Risk to the initial value

My system is not a good exemple du to the low win. Normaly, if your win ratio is greater than 50%, you should not decrease your risk$ when your account decrease. Your wining ratio should take care of your loss.

Better to stop trading for a given period if your account decrease by x% (you decide) until the market return to a normal pace for you.

Good luck
Martin

Exactly!

Nial did a case study of random entry and risk reward which you can read here -


http://www.learntotradethemarket.com...d-risk-reward/

I tried to play down lots when loss,,
but it's very hard,

increasing lot size after a loss is very risky
altought the chances for a winning trade are greater.
cause trader can lose all his money after series of consecutive losses.

here is a code(from LibDERKSUtils) that show
how to increase lot size after win
it determine the lot amount to use for the
"Dual Win" MM system.
If the last trade was a loss, then trade the same lots
If the last trade was a winner then check if it was preceeded by
loser(s) and adjust lot value if so.......

try it in your ea and see if it increases profit/DD ratio.

Sorry, post removed due to irrelevancy of the projected negative expectation (improbability) of the answer being of any benefit to either the thread starter in the long run and/or at least 99% of the viewers for the purpose of adding value to their lives.

Why not just risk a certain percentage every trade, like 1 % or 2. that way you account size will determine how much you risking every trade.