Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
Right, but isn't the loaded coin comparable to a strategy that does not generate profits?
I.e. if I used that loaded coin to perform my backtest I could see that by betting 100$ per toss I would be out of money in about 10-11 turns (on average): not a good strategy.
Also, regarding overtrading, I remember Mark Douglas talking about mechanical trading in his book and suggesting that it is not really up to you to decide when to enter the trade: you either have a signal or you don't, which means you either put a trade on or you don't. The example charts I provided above were all based on the presence of the price crossing the Bollinger Band. Shouldn't that take care of the overtrading bit?
Absolutely not! That coin game has a positive expectancy, no matter what you bet:
If you win 1 unit 90% of the time, and lose 1 unit 10% of the time, over the long term you will win 0.8 units per flip. Play long enough, and you'll have all the money in the world! That is why long term expectancy is important.
But, in the short run, you could hit a losing streak. And if your bet size was too big relative to your bankroll, you could get wiped out before you rule the world. That is why proper position sizing is important.
So, to win at this coin flip game, you have to be smart about the position sizing. If you are smart, you will win in the long run, practically guaranteed.
Not necessarily. a backtest with that coin shows only 1 possible outcome out of the infinite number of possible outcomes. So, you could easily have a backtest that never wipes you out. But the reality is that your future results could wipe you out, depending on how you bet and how the coin flips played out.
We are talking about 2 different things. I define overtrading as "too much size." To me overtrading has nothing to do with the strategy, as long as that strategy is long term profitable.
I agree with the late Mr. Douglas: if you develop a strategy, and it backtests well, you MUST take every trade that strategy says to take in the future, or else you are doing something different than the backtest. In that case, why even do the backtest?
So, first you develop a strategy with long term expectancy.
Then, you employ position sizing so you can maximize your gain without suffering too much drawdown, and minimize your chance of ruin.
Do not search for working trading systems that are being sold.
But search for some working systems here on futures.io that are for FREE.
You will find them!
I just finished a long and frustrating chart review today on exactly this point with a new guy working for me. Gaps all over the place in his charts and spreadsheets. I'm not blaming him, consistency takes stamina and this takes time. The least one can do however is to document every valid trade missed during a day. Randomising your (or donated) backtested system is only deceiving yourself going forward.
"A lie which is half a truth is ever the blackest of lies." Tennyson (backtest of lies is even worse).
What I meant to say was, a loaded coin with 90% probability is either hugely unfavourable or hugely favourable. It all depends on which side of the bet you are! (I am assuming I am always either calling heads or always calling tails)
I don't get this. If we are still talking about the backtest scenario with that loaded coin - provided we have a long enough sample period - and if I'm on the wrong side of the bet, shouldn't the backtest show it?
Note: I am assuming that being "on the wrong side of the bet" or "on the right side of the bet" means I always call heads or always call tails.
OK, that is true. I was confused because I assumed you'd pick the flip correctly, since you knew it was a loaded coin. If you always picked the wrong side, yes your equity curve would be down long term.
The problem with a backtest is that the data only deals with known items. However, the data going forward may be quite different. In the coin-toss example shown above, you would assume that over 10 bets, you would win 9. However, if you start keeping track of the actual results of 10 bets, you may find that winning 9 tends to happen more infrequently than it should. You may even have sets of 10 bets where you win none. You can run a Monte Carlo analysis which will highlight all of these different "roads" quite nicely.
To use William Eckhardt's example - assuming an unbiased coin (50/50 chance of heads or tails) is tossed an unlimited number of times, then the number of heads (or tails) you can get in a row approaches infinity. One would assume that such a large number of heads (or tails) in a row would never happen with that coin, but it still can.
Yes, it does go into it quite deeply. Have you ever considered how much of your results are due to luck? Some of my best periods were due to a combination of luck, ignorance and insane leverage, yet based on my results I would have looked like a damn good trader. However, I would not trade like that again - the risk of ruin was unacceptably high and even a short run of bad luck could have blown me up. In the same vein, even with a 90% win rate, you could still damage your account substantially if a long enough unlucky streak is hit.
Now, to get back to the current thread's topic, how does one determine whether results are due to skill, due to someone having solved a problem or due to luck? I would bet that the traders @DionysusToast mentions here
Of the traders that I know - they all rely on 'pattern recognition' to some extent but it is more related to behaviour than to a series of bars creating a pattern.
Trader 1 - Trades 40 or so stocks and has been doing …
have developed heuristics, i.e. rules that they apply in certain circumstances. These rules could potentially be programmed into a system and may make money.
If you were to look at successful traders and successful systems, you will find that their rules probably have a lot in common. I would suspect the thing that makes them successful is not the rules or the system, but rather several other factors. For instance, traders who risk too much tend to exit trades at exactly the worst moment (system traders may override stops in this case). Also, it takes a special kind of fortitude to trade through long drawdowns. How often do people change systems / rules once a losing period starts?
I don't think that the answer to the OP's post is black or white. There is a lot of gray (and luck) involved in how people make money (or lose money) in the markets.
Thanks grausch - a theme indeed seems to be emerging, which is that systems - if and when they work - are based on skillful/experienced implementation.
Being that the case, we should be able to conclude that trading is a skill, regardless of the approach (discretionary, systematic, or a mix of the two) that one adopts.
Practically there is a convergence from people from both camps towards experience and skill being paramount.
Do people agree, and would that answer @DionysusToast's original question?
The original question "Is trading a skill or is it a problem waiting to be solved?" is presented as an either/or proposition, and I think the answer is both.
You need skill to know how to solve the problem. That is true in any field, any endeavor. Take out the word trading in the original question and replace it with someone like "curing polio." That was a problem to be solved, but it took skill to do it (some mook off the street didn't find the cure, but a skilled research scientist did).
So in my mind successful trading is a problem to be solved, AND it takes skill to do it.