Intermittently … to be honest, I sometimes get exasperated at my own inability to explain this to people, and that may sometimes show in my posts. :8:
I have lost count of the number of occasions/threads over the last couple of years on which Ropunzel and I have each been involved in long conversations of which the key concept boils down to a misunderstanding of the “gambler’s fallacy”; and they usually end up with my abandoning the conversation in frustration, because eventually I re-learn - each time (until the next time) - that I’m apparently unable to explain it (even with a link) in a helpful way … the irony being that I like Ropunzel and want to try to help him, but I still fail [U]every single time[/U], and I know it, and it saddens me to see someone as intelligent and educated as he is apparently no further on towards becoming a profitable trader than at the time of any of the previous, essentially similar conversations we’ve had. :rolleyes:
Someone less stubborn (and Aspergerish) than I would probably simply shrug and keep out of these conversations, or just say “I’m really sorry but I just lack the didactic skills to help you with this”. :8:
Naturally, there is no sense in creating trading rules that cause stress or undue pressures, whether they are the result of losses or gains. One should alweays feel comfortable with one’s trading and flexibility is essential in that.
I think the situation you are describing (congratulations on that BTW ) is one of exceptional result beyond the original expectations, which is very different to the scenario of running a “normal” business on an ongoing basis. Of course, there is always good reason for pausing and taking stock of any situation.
But on an ongoing basis, it does not make sense, as mentioned above, to stop one day immediately when the target is reached but continue trading whenever losses are occuring? For example, on a day trading basis there are periods when markets are moving beautifully for weeks on end, but I remember the EURUSD last year in Sept/Oct when there was no follow-through on anything and one would only compound losses if one tried to trade out of those periods.
No not anecdotal evidence. I tested this with randomly generated trades, 10,000 outcomes. Starting at any point or any one of the trade outcomes and looking forward, the probability of ultimately having a higher account balance that you originally started with was 91%.
That’s not clear. Taking a win lose sequence (1,-1) on randomly generated trades with a 2.5:1 Reward: Risk. The equilibrium (Break even) strike rate is 28.57%.
So first of all starting at trade 1 - how many subsequent trades does it take for the win rate to be greater than 28.57%. I found that out of 10,000 starting positions i.e. start at trade 1,2,3…9000 etc…91% of the trade found themselves at a total strike rate above 28.57% that is the account balance would have been higher than the starting balance at some point 91% of the time. In some case it took just 1,2,3,4,5…200 subsequent trades to get there but finally it did and that is a random system with no edge.
The point is that with an edge system the upward movement of your account balance is not a linear one and is characterised by peaks an troughs. Setting a profit target is like having a non random walk with an absorbing win and lose barrier where the probability of moving toward your win barrier is higher than moving towards your lose one. If the Win barrier in 4% versus 10% on you Lose barrier and you move 1% up or down each step depending on the outcome of your trade with a higher probability of moving up each step (edge), you are more likely to even out your performance with the up barrier in place than without; hit up barrier reset.
This will make your trading outcome modular and predictable or quantifiable. In some case, depending on market conditions and the other factors that affect your performance; you will hit the win barrier in 4 trades (using this example) to continue trading will bring the possibility of making more, but also the possibility of making less and maybe even losing (at least in that time horizon - assuming you have a system with edge). It might be the case that it take 10,20 30 trades to hit the win barrier or you might find yourself never hitting it at all or even losing by hitting the lose barrier.
The ultimate result is that 1) even with a non edge system 50:50 system you will realise more consistent and predictable profit for an extended period of time, the difference being that ultimately, after 2,3,4 months your account balance will steadily decline. In a system with edge, your returns will be reliable and smoothed out for an extended period, with the occasion hiccup.
The EV of the system is itself a very subjective measure. With the original post in mind if I were calculate my win rate on Monday it would be 100% then would that be a suitable metric for working out the EV?
If you look a a distribution curve, be it bell, gaussian or poisson, the curve has a shape, hence the name curve, until after a large sample size it settles down to a figure in it’s “steady state”. That is what EV is like… forget the curve, it’s where the distribution settles in it’s steady state that matters for EV. (curve could be used to claculate risk of ruin so you can avoid it but lets not complicate things)
Forget short term, your win rate on Monday means nothing. It is impossible to know the sequence of wins and losses, no point in trying to do something that is 1) impossible and 2) unnecessary.
Win rate without expectancy doesn’t help much either.
I am pretty happy with my understanding of Gambler’s Fallacy and I am not saying that because I have a winning streak then a losing streak must follow. The point is I am wanting to modularise my trading outcomes to be more consistent. Part of the solution to this for me is setting a weekly target.
Now, on reaching my target early in the week; it may be the case that this week is good for me and I make even more money; great. It may also be the case that I lose more money and maybe even so much as to chalk the overall performance this week down as a loss.
I have hit my target; why take the risk! If I didn’t operate a system with Edge I wouldn’t be trading in the first place that is a galaxy away from the point. Even if I operate a system with edge (ultimately) does not guarantee that a positive EV will realise within the defined time frame i.e. with the required trade volume to hit my weekly target.
It may be the case that I start off badly in the week and have to put in 15 or more trades to finally get there, assuming there are even sufficient opportunities in the week that fit the parameters of my setups. If there are, and I’m confident in my system great, it will be done.
If it takes only 2 trades to get there then even better. Mine, yours, Goldman Sach’s system are not perfect. I set my target, I hit my target, there is no point exposing oneself to unnecessary risk: over trading.
If psychologically you find it difficult to be £4000 up on Tuesday and to see that drop by £3500 by Friday to £500 then that is one thing. The maths however is always going to be the maths, no way around it. [B]EV[/B]. The short term doesn’t matter.
I do however wish you good luck in your endeavours and hope that at some point you open up to the info being presented to you. I don’t think i can add anything more.
Sure but these curves assume an infinite number of trials whereas that is not the reality of the data that we work with. That is almost precisely the point, our data is noisy and finding a positive EV systems with a 1000 points of data does not tell you anything about the likely performance of your next 20 trades; they could be all loses and still even out to accretive EV.
That is the whole point, the next 20 trades losing do not matter if your money management parameters are correct and the system has a positive EV. You will make money in the long run. if you are not profitable now…try using EV.
Why the concern in the next 20 trades? Is your trading career going to be 20 trades long?
Ok thank you. You’re saying that if I trade more my profits will be higher assuming the EV is positive. Ok in return I say this to you. Pick a currency pair that you think is will close higher at the end of the year based on the fundamentals, then go long and don’t touch the trade till the end of the year.
Or would you try to buy low sell high each day/week/month. Tell me if you got it wrong and bough at a peak but where certain it would recover and move higher, what would you do close out at a loss or hold onto it, weather the downside.
A currency that will definitely close bullish by year end will nonetheless vacillate between peaks and troughs in that journey just as a system with positive EV will have good and bad spells. I am trying to catch the rising conscious of the fact that it may also fall. I would rather have consistent reliable yet smaller profits that take the full rollercoaster ride.
I understand your point clearly. I do not wish to join you in the 20 trade losing streak. Ill cap off at 4% you can do the 0.6% thing cheers. Next week we’ll start again.
Look at your trading performance week in week out for the last year, if you normally distribute the highest, average and lowest (negative), what is the probability that you will at least achieve a certain profit level at some point that week.
If i have an EV of £10: If trade 1) wins £150 to me it is just a £10 win . If trade 2 loses £100 to me again it is just a £10 win. It doesn’t matter if i have £2000 in profit on Wednesday or -£3000 it’s all irrelevant.
Each trade is a £10 win, the more trades the system provides the more £10’s i make. The final figure in “steady state” ie over the long term, will be the number of trades taken multiplied by £10.
[B]Above example for point proving purposes only…no compounding or “other” factors included… don’t try this at home folks[/B]
I have no idea the extent of your experience, if you are profitable or if you have a system with an edge but when multiple people who do this for a living are telling you the same thing…it just might be worth investigating.
If of course you are profitable and happy with the way things are working out for you then of course continue doing what works for you. It would be silly not to.
[I][U]Surely[/U][/I] you can see that that’s a logical fallacy, my friend? [B]In what way is your risk of trading on that week any more or less “[I]unnecessary[/I]” than if you stop for the week and start again the following Monday?![/B] :o
Sorry.
Enough of this. :23:
I regret that I’m unable to assist you in this matter, and I wish you well.
I honestly don’t get this.
If you’ve made your target, why stop and wait for next week/month to recommence?
You clearly have the right trading conditions for your system/set up now, why abandon it? Surely your trading rules still indicate that this is favourable?
Make hay while the sun shines, it could be months before you get favourable trading conditions again
The problem is that it is never that simple. You can not commodities the value of a single trade based on a system you have tested to be positive EV. Indeed if you have an edge and trade for long enough then ultimately your account balance will be higher for sure. But that tells you nothing about what will happen in a small trial set in the immediate future of a well define time horizon.
I can do a 1000 trade losing streak (0% strike rate) that evens out to a 55% strike rate overall after another 9000 trades. That system would technically have edge and therefore positive EV.
The EV you extract from the results of a system cannot be applied directly to you REAL expectation of your trades over a set number of trades or over a time horizon; as positive EV system only tells you that in the long run, for that particular pair in that particular test period the next outcome of wins and loses was an overall net win.
You cannot apply it directly, because you cannot predict accurately the exact probability of a trade being successful at any given time; I cannot build a system on EURUSD with data for 2014 - 2016 that proved to have profitable (i.e. positive EV) with a win rate of 53% say on a 1:1 R:R trade and then therefore conclude that every trade I take with than system on that pair has a 53% probability of winning.
There are underlying market factors that determine on a granular level, if my strategy is more or less successful at any given time, it simply averaged out to 53% over the data set I chose to investigate.
What I do know for certain is that across multiple currency pairs in a defined period of time there are likely to be a combination of good and bad spells. Infact I can probably investigate the PnL on my account every given week and predict with 90% certainty that at some point in time in that week, my PnL might be at least say 2% up.
Why wouldn’t I then, to regularize my performance, choose to cap of my trading at that point; yes I might opportunites, but more worryingly I might lose more money. At least that way there is some peace of mind and less erraticism in the trading outcomes.
I can look at the market now and start crying about all the opportunities I missed today, Indeed I could have turned 10,000 into a million bucks just today if I made all the right calls, but is that realistic.
Time. My aim is hit a weekly target. @ 4% return on monday. The probability that hit my weekly target is 100%. If I continue to trade, there is chance that I make more; great, or lose as was actually the case.
If I am down -3% on Thursday, do you suppose the probability of getting to +4% by Friday end in the number of available trades matching my setups (number of available steps) is the same as being 0% on Monday Morning.
4% is banked, Now. There is a time value in realising that money now.
Maybe I’m crazy. I think it’s a good idea to have a period specific profit goal in mind. Like I have said many times even with ‘flawless’ edge you are still exposing yourself to loss every time you pull the trigger on a new trade.
I hit above my weekly target midday Monday with 4% return and 4 days to go but foolishly continued to trade to arrive at 2% wednesday evening, giving only 2 days to recover those losses; there was no point doing that.
I would favour certainty over greed. I am talking specifically about intraday volatility trading with technical analysis, with a weekly target it makes more sense to settle as close to your target as possible as you edge closer to Friday market close.
Yes you can make more money in the week but you can also lose more too and I would prefer to be extra vigilant about that possibility. Next week would afford me 5 full days to try to achieve the same target.
Even if it took 15 trades to get close to the target next week, a 100% win rate this week would reflect positively on my system’s strike rate and therefore EV; indeed to give an exaggerated example to make my point. if every week I started with 2 win trades followed by a medley of wins and losses to arrive at 0.6% for the week; would I not have a better performance, strike rate and EV if I quit after the first 2 trades each week.
) is one of exceptional result beyond the original expectations, which is very different to the scenario of running a “normal” business on an ongoing basis. Of course, there is always good reason for pausing and taking stock of any situation.