after a number of trades from the same strategy, why my winners win? why my losers lose?
- Thread starter tomdwan
- Start date
To me - my own interpretation, is that de facto your trade entries and your position management were also random ? If not, please elaborate.
The “roll of the dice” statistical exercise is literally Junior High math.
If you can't explain why your winners...win -- and your losers...lose...you shouldn't be trading, at all. Because you got nothing.
Winners make it happen...they carve and view and approach the market game to their advantage and skill.
While losers rely on prayer and voodoo and luck and wishful-thinking and mysticism.
If you're trading based on completely randomness, blindfolded, and just letting the numbers essentially be more positive then negative in the end...you're not trading, but gambling. and needless to say, destined for inevitable doom.
Everyone in the casino loses in the long run. Don't be one of those silly people who short-term wins and thinks they are a genius.
Last edited:
Blame your losers on volatility.You`ll never predict volatility, hence you`ll never define your stops straight.Apply Game Theory to your strategies and then maybe you`ll have a chance.Roll the dice.
You had 400 losers because your strategy built them in, it has a 60% win rate, the 40% of losers is not avoidable except over a very short, statistically and financially insignificant run of samples.
There is a danger that you will start to think either the 40% of losers are down to your personal errors of judgement, or you can tweak things to ramp up the win rate. Both reactions are natural but wrong.
The only outcome you can possibly change significantly is the amount of capital you risk and win per trade. There might be ways of cutting some of your losers earlier and increasing the win size on your winners.
Well, he may possibly (probably?) be able to increase the win-rate by taking the same entries but changing the proportional relationship between the targets and the stop-losses, but it would be a mistake to imagine that increasing the win-rate in that way would also necessarily increase the profit. (I'm being pedantic - sorry! - but this is one of the things you have to test, of course ...
Well, he may possibly (probably?) be able to increase the win-rate by taking the same entries but changing the proportional relationship between the targets and the stop-losses, but it would be a mistake to imagine that increasing the win-rate in that way would also necessarily increase the profit. (I'm being pedantic - sorry! - but this is one of the things you have to test, of course ...).
This is certainly so, and I think we agree that as you change one parameter in a trading strategy, it will affect others - sometimes in unforeseen ways. Yes, win rate might be increased as a side effect if you like of other tweaks, but should never be an objective. Win rate increase is probably going to be in single digits of %, but its possible the increase in profitability, which is the real goal of course, could be far more important.
%%You're joking (I hope)?! Over 1,000 outcomes, it's more than 6 standard deviations from the mean: they don't often come much more conclusively non-random than that ...
Sounds like TomD may on to something good. I would not call it random, not if he tests it for 2,000 days.1000 could be to close to random. Frankly i hate the word random; non random is fine.It also appears he is not a snake oil salesman.LOL Just improve it; if that system loses 400 or 800 times in a row.Of course 400 losses/667 gross winners first year does not mean much @ all.Its not like gambling in pool halls @ all; good question.
well, unlike the majority of members who have replied to this thread, i would never use ordinary statistical analysis to evaluate a trading strategy. prices for financial instruments are really time series not variables subject to any statistical distribution; they have one and only one value for each moment in time and it is never possible to repeat an observation as each moment in time happens only once and is then gone forever. ordinary statistical analysis is appropriate for stuff like throwing dice and other events that can be repeated as many times as necessary and the resulting observations can, under strictly controlled conditions, credibly be considered identical phenomena to one another. additionally, i would never consider price action of publicly traded financial instruments to be anything close to a random occurrence in this rigged world where stupid central banks have injected more than the equivalent of 20,000,000,000 usd into financial markets since 2009. all that stupid "liquidity" plus the manipulation of futures contracts is done to make sure that prices of financial instruments only move in a very predictable, upward, rigged trajectory. the bank of nippon currently owns more than 75% of publicly traded etf's in the markets of nippon and recently stated it would buy an infinite number of nippon bonds at something like 1.1%; so, anyone who pretends to consider prices of financial instruments as random variables should be laughed at and disposed of.
now, if and when you are able to develop a trading strategy that operates on rigid, fixed rules (such that could be put into computer code and automated) and you want to evaluate the results it generates, my interpretation would be the following: any strategy is just a set of logical rules that follows the price of a financial instrument waiting for particular price patterns - price action to trigger entries, position management and exits. if any particular strategy on any particular instrument generates 600 hypothetical positive results that would just mean that the price of the instrument completed some particular pattern and it then followed through as the strategy expected which led to a positive result. same with the 400 hypothetical negative results, the price of the instrument initially completed some particular pattern but then didn't follow through in a way that the strategy could benefit from, which led to negative results. that's all.
i have developed a number of automated strategies; they have positive results when price moves in patterns favorable to my strategies, don't trade if the setups i'm looking for are not there (absence of significant price movement) and generate losing trades when the setup is initially triggered but then price action moves in a way that my strategy cannot profit from. that's it. i will give two examples for illustration. cl and ng, had very large average movements every day and every month up to 2009, after that date, their movements have consistently been still large but far more subdued compared to the previous period. any strategy that generated profits up to 2009 on those instruments would not have had the same kind of results after that date because those instruments move in very different ways since. the rules for any hypothetical strategy would be the same and they would have worked before 2009 but would not work after that date because those instruments don't display the same patterns anymore. it's the same story with stocks that have split, if you had strategies that made money on let's say nflx or aapl before their latest splits, the very significant change in scale after the split will mean that those strategies won't generate the same kind of profits they did before the split. and the same situation applies in reverse for strategies refined for the price action that has been observed after the split if they were to be evaluated on the data before the split.
Trading is a game of probabilities.
We all would like all our trades to be winners, but we know this is not possible. We know some of the trades will be losers. Many traders think that if a trade has lost money, it was a bad trade. They try to identify what errors they made that lead to losses. Why? "Because I lost money! So surely I have made a mistake somewhere?”
Was it the right conclusion? Is any losing trade necessarily a bad trade?
The answer is no. No matter how well he executed his trade, there will be losing trades because we are playing a probability game. Trading is a business based on probability. And probability means that sometimes we get what we want, sometimes we don't. And that's the nature of the this business. The sooner we accept this, the better we can operate it as business.
"There's a difference between knowing the path... and walking the path." - Morpheus
Example: if you trade options and implement a strategy like iron condor that is based on high probability of the underlying remaining between your short strikes, you might win 70-80%, but you will still lose 20-30%. This is completely normal.
We all would like all our trades to be winners, but we know this is not possible. We know some of the trades will be losers. Many traders think that if a trade has lost money, it was a bad trade. They try to identify what errors they made that lead to losses. Why? "Because I lost money! So surely I have made a mistake somewhere?”
Was it the right conclusion? Is any losing trade necessarily a bad trade?
The answer is no. No matter how well he executed his trade, there will be losing trades because we are playing a probability game. Trading is a business based on probability. And probability means that sometimes we get what we want, sometimes we don't. And that's the nature of the this business. The sooner we accept this, the better we can operate it as business.
"There's a difference between knowing the path... and walking the path." - Morpheus
Example: if you trade options and implement a strategy like iron condor that is based on high probability of the underlying remaining between your short strikes, you might win 70-80%, but you will still lose 20-30%. This is completely normal.