Do you use bet size to compensate for losing trades?

[u0362565]

Hi all,

I was wondering how others view the bet size? in trades.  I'm talking about the £/point figure.  Assuming most go for high reward/risk ratio that means you don't need such a high win rate however e.g. a win rate of 33% may seem high to some for a 2:1 ratio.  So assuming you can't achieve 33% you can't keep plugging away with the same bet size expecting to achieve profit.  So do you make your calculations for the next trade considering that you need to make up for prior losses? E.g. if you've lost more than 3/10 trades.  I find it hard to believe that people can achieve a consistent win rate that doesn't involve increasing the bet size to compensate for losses.  Or do you always use one figure e.g. £5/point.  Perhaps this is the rub, you can keep the bet size the same if you truly can achieve said win rate consistently.  For those who can't, give up? Alternatively don't adjust bet size but adjust the position size, that's another option but keeping the ratio the same.

[HMB]

13 minutes ago, u0362565 said:

Hi all,

I was wondering how others view the bet size? in trades.  I'm talking about the £/point figure.  Assuming most go for high reward/risk ratio that means you don't need such a high win rate however e.g. a win rate of 33% may seem high to some for a 2:1 ratio.  So assuming you can't achieve 33% you can't keep plugging away with the same bet size expecting to achieve profit.  So do you make your calculations for the next trade considering that you need to make up for prior losses? E.g. if you've lost more than 3/10 trades.  I find it hard to believe that people can achieve a consistent win rate that doesn't involve increasing the bet size to compensate for losses.  Or do you always use one figure e.g. £5/point.  Perhaps this is the rub, you can keep the bet size the same if you truly can achieve said win rate consistently.  For those who can't, give up? Alternatively don't adjust bet size but adjust the position size, that's another option but keeping the ratio the same.

I did. It led to catastrophe.  Read my 2nd post on this site.  If I could start again, I'd NEVER increase size as a reaction to a loss.

[HMB]

This is the only  thing I know without doubt about trading:  trying to make up for losses with larger trades is the way to ruin with certainty.

Read you own post again and ask what it would imply:  if you can't make money with a small risk, you take a larger risk.

And what if that goes wrong?  Of course - double up.  And then it fails again..?  Then you're walking to the bank to ask for a consumer loan...

Please read that "cold" warning about 76% of traders "encountering losses" again - and try to add some imagination to it.  Try to imagine ruin with all its consequences.  And then ask yourself how people got to that point. 

Edited October 20, 2020 by HMB

[u0362565]

Yes i see the problem for sure, but one thing is true, it does allow you potentially many goes potentially far more than is statistically likely for you to lose.  So would negate the requirement for a high win rate, however i see that if you do exhaust your number of attempts it becomes like having a very large risk for not much reward, but it is about the probability in the end.. Its if you can stomach it.  Personally i struggled and if you think it could affect your health is it worth it, no.  Surely most traders are stressed most of the time!?

[HMB]

Just now, u0362565 said:

Yes i see the problem for sure, but one thing is true, it does allow you potentially many goes potentially far more than is statistically likely for you to lose.  So would negate the requirement for a high win rate, however i see that if you do exhaust your number of attempts it becomes like having a very large risk for not much reward, but it is about the probability in the end.. Its if you can stomach it.  Personally i struggled and if you think it could affect your health is it worth it, no.  Surely most traders are stressed most of the time!?

I will not engage in this discussion further.  This is too much responsibility for me, sorry.  You should discuss this with someone who truly knows you - I suggest your parents or similar.  Not joking, no matter your age.

Also speak with experienced traders and try to listen.  But you likely won't.  And in ten years you will likely write a similar response as I'm doing now.

 

DON'T INCREASE SIZE TO MAKE UP FOR LOSSES - THIS IS THE MOST DANGEROUS MISTAKE YOU CAN MAKE!

 

It's about letting profitable trades run, and cutting losses early, so that you are with a low win rate still profitable 

 

I hope for others in this forum to support me here.  But I'm out. 

 

THIS IS NOT FINANCIAL ADVICE.

 

[TheGuru12]

Yeah, Never go chasing losses. 

[THT]

NOOOOOOOOOOOOOOOOO

This is how I view Risk - If my account value is £x then the max per trade I'm prepared to lose is 2% of the £x - lets say the £x = £2,000

The the most I'm risking is £40 per trade I take (The £ value of this 2% changes as my account goes up or down) 

Then on the market I'm trading for ease, say Entry = 100p and I ascertain that my stop needs to be at 90p = 10p range of risk

Then the £ per point is simply = £40 / 10p = £4 per point

Read Dr Van Tharps - Trade your way to Financial Freedom - it explains the laws of probability and probabilistic returns - these laws CONTOL things when you trade, they cannot be avoided

You need to build into your mind a mechanical emotionless setting everytime you trade - it should not matter whether you win or lose, the way I do this is I take the trade happy to take a loss

Depends on your method too IF (and most traders don't know this) you know the exact stats of your method over many many trades then you could just trade a fixed £ risk amount and know that at some point it'll come good if you take a series of losses one after the other

 

[DSchenk]

Isn't this the old roulette "strategy" where you bet on red or black.

Bet $1, lose $1.

Bet $2, lose $2. Total Loss: $3

Bet $4, lose $4. Total Loss: $7

....

Bet $256, Lose $256. Total Loss: $511

Bet $512. Win $512. Total Win: $1  

Works 100%, except when it doesn't and red comes out 50 times in a row, whilst you are on black  

Edited October 20, 2020 by DSchenk

[jlz]

Martingalers at it again, what could go wrong?  

Listen to HMB and THT, their words are coming from experience. 

[HeOfMuchDisgrunt]

Have you (the OP)  not heard of the Kelly criterion? Lots of info on the internet about it, including versions applicable to financial trading. The Sharpe Ratio is a related idea.

The basic premise of Kelly  is that you risk a percentage of your bank that is proportional in some way to your best estimate of your edge. The more certain the trade the higher the amount. Nevertheless after a series of losses your actual bet size is reduced for a given situation no matter how confident you might be. Say Kelly says bet 2% of your bank on one especially good trade (meaning set the position size and stop-loss so that if it gets hit you lose at most the whole 2%). If you do multiple bets in succession that all lose then that 2% of your new, say, half-bank is actually only 1% of your original bank. This way you can avoid going bust. It can be proved mathematically that with a given known edge the Kelly method will grow your money the fastest, but it will also have rather volatile swings, doubling and halving not being at all unusual. Many pro-gamblers and traders use a fractional Kelly amount to get most of the gain but less of this undesirable level of volatility - often only a quarter to a third of the Kelly amount. This allows some leeway in the accuracy of the estimate of your edge. If you use Kelly but bet higher amounts than the amount recommended by the formula you can actually still go bust even if you have a large and precisely-valued edge, so it's far better to underestimate your edge and bet too small. The goal will take longer but at least you have a good chance of getting there instead of busting out.

[u0362565]

So i guess the answer is no you don't adjust bet size to account for losers.  Good to know thanks.  Quite astonishing that anyone makes any money out of this game, hats off to you.

[u0362565]

Sorry game was the wrong choice of words..

[andysinclair]

2 hours ago, HMB said:

t's about letting profitable trades run, and cutting losses early, so that you are with a low win rate still profitable 

This is the key for me.

It doesn't matter if you have more losers than winners, what really matters is how you manage your risk.

[HMB]

5 hours ago, dmedin said:

Take your tongue out of other people's bottoms, little Swedish hacker retard.

this probably should have been ignored, but IMHO it was a new low.  obviously jlz tried to help someone here, and referred to others to add emphasis to a point.  I'm writing this because I actually asked for support in my earlier response, and I'm very pleased I'm not the only one who reacted to this post thoughtfully, so I think it would be pretty weak if I didn't speak up here.

That may make me look ridiculous, and like my tongue is now travelling through a bunch of cracks, but I don't give a **** (like I think you would not either).   

I'm actually surprised that you @dmedin  left it to this statement - usually you're significantly more effective in warning people of the hazards we face as retail punters IMHO - I think you definitely had stronger moments.

you know what I would find useful:  having a separate, focused thread for insults  - I'm sure by now the respective targets wouldn't mind (to not again say not give a ****), and other discussions could become - let's say "leaner".

I genuinely think I got a lot of valuable insights from many of your posts here, and I highly appreciate the honest feedback, and I also believe you have good intentions. I don't know about any history between you, jlz, THT, and Caseynotes.  I further think you play an important role here in this forum and I look forward to more productive discussions with you.

But the post quoted above was a new low.   

 

[jlz]

9 hours ago, HMB said:

this probably should have been ignored....

That is why I wrote the Javascript script I posted the other day. I don't see his posts. I went to look for the post just to see the insults and, yes, they are quite low, even for him.

This forum doesn't have the option to ignore a user completely. On other forums when you ignore a user that person dissapears but on this one you still see the user unfortunately. They can tag you and make you read their nonsense.

The main reason why he is insulting me is because I exposed him. I exposed his b*llshit so clearly that he can't cover it up. As a result he is only not insulting me but everyone around. 

It is just a waste of time, don't worry at all. I don't see the point of replying to him at all.  

[dmedin]

10 hours ago, HMB said:

I'm actually surprised that you @dmedin  left it to this statement - usually you're significantly more effective in warning people of the hazards we face as retail punters IMHO - I think you definitely had stronger moments.

I know right, it's my Achilles heel, pretty much destroyed my life by indulging in foul-mouthed outbursts.

[Caseynotes]

10 hours ago, HMB said:

this probably should have been ignored, but IMHO it was a new low.  obviously jlz tried to help someone here, and referred to others to add emphasis to a point.  I'm writing this because I actually asked for support in my earlier response, and I'm very pleased I'm not the only one who reacted to this post thoughtfully, so I think it would be pretty weak if I didn't speak up here.

That may make me look ridiculous, and like my tongue is now travelling through a bunch of cracks, but I don't give a **** (like I think you would not either).   

I'm actually surprised that you @dmedin  left it to this statement - usually you're significantly more effective in warning people of the hazards we face as retail punters IMHO - I think you definitely had stronger moments.

you know what I would find useful:  having a separate, focused thread for insults  - I'm sure by now the respective targets wouldn't mind (to not again say not give a ****), and other discussions could become - let's say "leaner".

I genuinely think I got a lot of valuable insights from many of your posts here, and I highly appreciate the honest feedback, and I also believe you have good intentions. I don't know about any history between you, jlz, THT, and Caseynotes.  I further think you play an important role here in this forum and I look forward to more productive discussions with you.

But the post quoted above was a new low.   

 

are you mad? none of this is new, it's straight abuse by a moronic attention seeker and systematic malicious acts to derail and destroy threads where people have worked hard for the benefit of others, been going on for months, there is no valued contribution, it's simple total negativity by someone with obvious mental health problems and is not only of no use to anyone else but is detrimental to all.

Obviously at some stage the moderator will actually have to do something, not a problem for me I can wait but the forum suffers in the meantime.

[dmedin]

15 minutes ago, Caseynotes said:

are you mad? none of this is new, it's straight abuse by a moronic attention seeker and systematic malicious acts to derail and destroy threads where people have worked hard for the benefit of others, been going on for months, there is no valued contribution, it's simple total negativity by someone with obvious mental health problems and is not only of no use to anyone else but is detrimental to all.

Obviously at some stage the moderator will actually have to do something, not a problem for me I can wait but the forum suffers in the meantime.

 

You are a miserable old b4stard.  Go and post more sh!t about 'globalists' in your Covid-obsessive thread.

I hope you enjoy the Biden-Harris victory in two weeks as well.  Trump belongs in jail 🤣

[u0362565]

21 hours ago, HeOfMuchDisgrunt said:

Have you (the OP)  not heard of the Kelly criterion? Lots of info on the internet about it, including versions applicable to financial trading. The Sharpe Ratio is a related idea.

The basic premise of Kelly  is that you risk a percentage of your bank that is proportional in some way to your best estimate of your edge. The more certain the trade the higher the amount. Nevertheless after a series of losses your actual bet size is reduced for a given situation no matter how confident you might be. Say Kelly says bet 2% of your bank on one especially good trade (meaning set the position size and stop-loss so that if it gets hit you lose at most the whole 2%). If you do multiple bets in succession that all lose then that 2% of your new, say, half-bank is actually only 1% of your original bank. This way you can avoid going bust. It can be proved mathematically that with a given known edge the Kelly method will grow your money the fastest, but it will also have rather volatile swings, doubling and halving not being at all unusual. Many pro-gamblers and traders use a fractional Kelly amount to get most of the gain but less of this undesirable level of volatility - often only a quarter to a third of the Kelly amount. This allows some leeway in the accuracy of the estimate of your edge. If you use Kelly but bet higher amounts than the amount recommended by the formula you can actually still go bust even if you have a large and precisely-valued edge, so it's far better to underestimate your edge and bet too small. The goal will take longer but at least you have a good chance of getting there instead of busting out.

No not heard of it, will check it out thanks for the info.  I guess i would take issue with the concept of "the more certain the trade".  I find in trading nothing is certain so I can't put more money on one trade vs another. 

[HeOfMuchDisgrunt]

If you would take issue with the concept of "the more certain the trade" then you have answered your own question. When you have had a run of losers then, absent any other evidence, you have less certainty therefore if you had been using a fixed proportion of your bank per trade you would now have reason to reduce that proportion until your winning trades became more numerous. The Kelly formula for a 50-50 bet with equal win/loss amounts tells you to invest a maximum amount of zero in the bet. In other words don't play the game at all if you have no positive expectation. Your suggested solution to a losing streak was to increase your bet-sizing, yet you say you have an issue putting more on one trade than another. That's a contradiction. If you cant judge individual bets you can at least look at your recent results - poor results suggests a smaller edge (or no edge at all), which leads to smaller bets or staying out of the market altogether - this is the complete opposite of the idea put forward in your opening post.

Kelly says to bet smaller or not at all until you have a level of certainty on a trade that is greater than 50%, whereas your idea was to increase the bet size when your recent results suggest a level of certainty somewhat less than 50% - one method will preserve your capital until you find trades you are more confident about, whilst the other could lead to you going broke. Your choice.

 

[u0362565]

21 hours ago, HeOfMuchDisgrunt said:

If you would take issue with the concept of "the more certain the trade" then you have answered your own question. When you have had a run of losers then, absent any other evidence, you have less certainty therefore if you had been using a fixed proportion of your bank per trade you would now have reason to reduce that proportion until your winning trades became more numerous. The Kelly formula for a 50-50 bet with equal win/loss amounts tells you to invest a maximum amount of zero in the bet. In other words don't play the game at all if you have no positive expectation. Your suggested solution to a losing streak was to increase your bet-sizing, yet you say you have an issue putting more on one trade than another. That's a contradiction. If you cant judge individual bets you can at least look at your recent results - poor results suggests a smaller edge (or no edge at all), which leads to smaller bets or staying out of the market altogether - this is the complete opposite of the idea put forward in your opening post.

Kelly says to bet smaller or not at all until you have a level of certainty on a trade that is greater than 50%, whereas your idea was to increase the bet size when your recent results suggest a level of certainty somewhat less than 50% - one method will preserve your capital until you find trades you are more confident about, whilst the other could lead to you going broke. Your choice.

 

Yes for me at least i have no certainty in a trade, don't know how anyone does really.  I don't really have a point i was just thinking about what tools do the platforms actually give you, indicators yes, stops, limits and bet size.  There is so much variation in the markets that i thought that bet size may play a key role because its one of the only things you can vary.. I just don't understand with any risk/reward ratio how the win rates can be achieved frequently enough that any account increases without playing with bet size as a means to give you more attempts.  I guess this is why trading is so difficult, a lot of websites go on about if you have a high risk reward ratio you only need a 20-30% win rate and at first you're like that's not so bad but then when you have a go actually its very tough to achieve consistently.   

[HeOfMuchDisgrunt]

You could have a system that gave you 90% winning trades and still lose if your bet-sizing is too big. 

I think you misunderstand my use of the word "certainty". A coin-toss has no bias, therefore your certainty of winning a series of tosses in the long run is zero. If your own normalised historic record of picking winning trades and cutting losers leaves you with some level of consistent average profit per unit-sized risk then you could say you have a degree of certainty in your trades. Or perhaps "confidence" would be a better word to use.  The more confident the bigger the bet-size, but don't bet more than a few percent on any one trade - if one bet loses 2% of your bank then you need to make slightly over 2% of bank over your next few trades to get back to break-even but if one bet loses you 50% of your bank then your next few bets would have to increase your bank by 100% just to get back to where you started - that's very hard to achieve! To do it properly at a 50%  bank your bets that were only 2% of your original bank are now 4% of your current bank so you have to cut the size in half, meaning the recovery will take a long time - all because of one stupid over-sized bet that went wrong. 

Note:- reducing your bet-size means betting a smaller amount per trade, not simply betting the same per-point amount but narrowing your stop, cos narrow stops will get hit very often. Remember that your actual bet size is your stop distance times your per-point bet.

[u0362565]

On 06/11/2020 at 19:46, HeOfMuchDisgrunt said:

You could have a system that gave you 90% winning trades and still lose if your bet-sizing is too big. 

I think you misunderstand my use of the word "certainty". A coin-toss has no bias, therefore your certainty of winning a series of tosses in the long run is zero. If your own normalised historic record of picking winning trades and cutting losers leaves you with some level of consistent average profit per unit-sized risk then you could say you have a degree of certainty in your trades. Or perhaps "confidence" would be a better word to use.  The more confident the bigger the bet-size, but don't bet more than a few percent on any one trade - if one bet loses 2% of your bank then you need to make slightly over 2% of bank over your next few trades to get back to break-even but if one bet loses you 50% of your bank then your next few bets would have to increase your bank by 100% just to get back to where you started - that's very hard to achieve! To do it properly at a 50%  bank your bets that were only 2% of your original bank are now 4% of your current bank so you have to cut the size in half, meaning the recovery will take a long time - all because of one stupid over-sized bet that went wrong. 

Note:- reducing your bet-size means betting a smaller amount per trade, not simply betting the same per-point amount but narrowing your stop, cos narrow stops will get hit very often. Remember that your actual bet size is your stop distance times your per-point bet.

Thanks for that. I will bear all this in mind in future. I can see that increasing bet size is not the way to go.