Is this a good system long term?

I've been back testing my long-only system for the past 45 trades (stocks, not fx), and I've noticed that on average my Risk/reward ratio is not good at all, just .55, however I'm up 39% on the back tested trades with 64% winners.

Should a trading system be focused on RR, or can a system with high positive expectancy ignore the RR ratio all together?

thanks

bump

bump!!

Hi,

I don't mean to be pessimistic, but 45 trades seems a little too small to be of any use (statisticaly).

The fact that you have 64% winners could be explained by the fact that you ran into a short term trend allowing your system to work (and later on, might blow your account).

However, if on a greater sample, the % are confirmed, then have fun trading your system sir

The trades are across many different stocks. But yes, it is a trend following system.

Yes long term strategy is good in a sense that we make much profits in it. Traders wait for specific time of trade when they find their favorite market conditions they act upon their plan .Long term traders are much experienced and active they get their targets in only some trades

if you're ready to withstand big dropdowns, and have a lot of patience (when your orders in not bad plus... Why don't you close? ) I prefer big lots and fast profit

Regarding 0.55 Risk/reward ratio being "not good at all, just .55" since it is Risk over Reward, having a low number is good. If you flip it over for Reward/Risk you get 1.8 profit for every 1.0 of loss - this is essentially like profit target being 1.8 times further away than stop loss. Combined with your high win % (above 50%) and you have a decent looking system with positive expectancy.

But as Akh12851 pointed out, you have too few trades to have any kind of confidence that performance will continue with similar win% in the future. I did an analysis for the stats you reported using code for computing confidence intervals for win % and came up with:

90% confidence interval range: 0.23
"lower conf bound: 0.52 upper confidence bound: 0.75"

So with 90% confidence with only 45 trades you could say that your true win % could be as low as 52% and as high as 75%. Given the 1.8 to 1 Reward to risk ratio it should still be a profitable system with 90% confidence.

99% confidence interval range: 0.35
"lower conf bound: 0.45 upper confidence bound: 0.8"

With 99% confidence you could say that your true win% could be between 45% and 80%. Even at 45%, with a 1.8 reward/risk ratio this system should be profitable.

Congrats, big_pipin, assuming your system hasn't overfit the data and the system generalizes well on new data, it looks like you've found something interesting.

No the system should not be focused on RR. Most important thing in every system is MM. The other parts of the system can be variable. My opinion is that 1:1 RR is good enough if you are confident in your knowledge and edge. The market is not random and sometimes you just have to close your trade instead of waiting like a dummy because your RR is 1:3

I completely agree with kvv888 on the part that the most important is Money Management.

Let's take the R:R example :

I use 1:3 RR :
1st Trade : opportunity for a TP of 90 pips => SL placed at 30 pips
2nd Trade : opportunity for a TP of 270 pips => SL placed at 90 pips

On the paper, all seem good. now, let's say :
- 1st Trade was a winner, 2nd was a looser => Profit = 0 pips
- 1st Trade was a looser, 2nd is a winner => Profit = 240 pips

Over 2 trades, with a fixed RR and let's assume a 50% winrate, results can change greatly depending on the outcome of your trades. And your RR will not save anything in such a case.

In my opinion, RR should only be used as a tool for studying the past (in the end of the year, it is nice to know that you risked x$ in order to make a profit of y$), and in no case to anticipate the future.
There is another case in which you can use observed RR in order to make assumption for the future : RR is fixed, and so are TP and SL (always going for x pips TP and y pips SL)

Have all a great day,

If you have 64% winners, and a risk-to-reward ratio of 0.55, then after 100 trades, on average your figures should look something like this ...

64 winning trades winning you 1 unit each = +64 units
36 losing trades costing you 0.55 units each = -19.8 units
Balance over 100 trades: + 44.2 units
Profit Factor (64/19.8) = 3.23

It's hugely profitable, and looks like a dramatically successful system. But 45 trades isn't a statistically significant number of trades, so it's too early to tell this. And backtesting isn't the same as forward-testing (for many different reasons). When you've forward-tested 200 trades, if your results are somewhere near that, then you probably have a complete triumph on your hands.

Alternatively, if by some mischance you meant reward-to-risk ratio rather than risk-to-reward ratio, then the situation is radically different, because after 100 trades, on average your results will look like this ...

64 winning trades winning you 0.55 units each = +35.2 units
36 losing trades costing you 1 unit each = -36 units
Balance over 100 trades: -0.8 units
Profit Factor (35.2/36) = 0.98 (PF < 1 shows that the system loses money).

In this forum, in particular, members are sometimes prone to discuss what they refer to as "RR ratios" without specifying which way round they're calculating it. As you can see, it can very easily make the difference between huge profit and breakeven or loss.

You see the difference, doubtless? A risk-to-reward ratio of 0.55 means that your average loss is 0.55 units and your average win is 1 unit. A reward-to-risk ratio of 0.55 means that your average loss is 1 unit and your average win is 0.55 units.


Not "altogether"; no.

In general, experienced and successful traders tend, overall, to want to avoid very high win-rate systems.

These are typically systems which regularly make small profits but incur occasional disproportionately huge, disastrous losses. Such systems, unfortunately, tend to be both easy to research/discover and easy to promote/publicise because they appeal to very inexperienced, unsuccessful traders who mistakenly imagine that "win-rates" are the most important aspect of system selection.

If you read an introductory trading book such as Van K Tharp's Trade Your Way to Financial Freedom or Tushar S. Chande's Beyond Technical Analysis (both very strongly recommended), you'll find plenty of detailed explanations (with examples) of why it's so much easier and more plausible for people - especially beginning/inexperienced traders - to make profits from systems/methods with much lower win-rates than higher ones, and why seeking out systems with very high win-rates usually ends up being such a big mistake.

Hope it helps - and good luck!