How to find risk reward in your investment?

It’s pretty simple. Look at the industry standard sharpe ratio. Higher number means higher returns per dollars. Sharpe ratio 1.0 or above is recommended.
Steps to find Sharpe Ratio in the selected strategies.
Step 1.

Step 2:

Check it out this important article.^dsa-1427141793820^b^547854587984^^^g

Sharp ratio is a totally meaningless metric, it cannot distinguish between upside and downside volatility.

In other words, if your winning trades are much bigger than your losing trades (in absolute value), the sharp ratio will say that your system is too “volatile” and “risky”, which is total nonsense of course.

A much, much better metric is the adjusted Sortino ratio.


May be you should comment on authors blog. If someone carefully evaluate all the ratios then he can have better idea about strategy.

I could not find Return Over Maximum Drawdown (RoMaD) on C2. It’s very interesting parameter to evaluate risk reward.

1 Like

This is indeed a very valuable metric I would love to see on Collective2.

1 Like

When all academic researchers want higher Sharpe Ratio, you said SR was meaningless :joy:

This is Calmar Ratio

Academic “researchers” are the worst traders in the world, and most of them have no clue about real life trading.

For example Long-Term Capital Management (the now infamous hedge fund) had not one but two Nobel Prize winners in Economic Sciences in its team, and yet the fund totally collapsed and then disappeared when it lost close to $5 BILLIONS in 1998, thanks to these “brilliant” geniuses.

For C2 trend-following systems, the sharp ratio is totally useless.

Not exactly, the Calmar ratio uses the risk-free rate in its formula, while here we are only concerned about the return to drawdown ratio (also called the Pain to Gain ratio).

Classic Calmar ratio is the ratio of the annual return to max dd over three years. Nothing else. Sharpe has risk free returns in it.

Here is the formula for the Calmar ratio:

We always assume the risk-free ratio is 0% because the SR can change when you change the strategy’s overall leverage level.

And you basically cherry-pick the best example to explain why academic researchers are the worst. I really disagree with this point. Despite most researchers are not good and their research papers don’t have money value, you should look at those good researchers (eg. NBER researchers, Berekely Lab researchers).
What they are doing is science and their approach is data-driven. This means they understand more than normal people averagely.

BTW, most ridiculous point is that you said SR was useless… :joy: :joy:

For the few profitable trend-following systems, the Sharpe ratio is absolutely worthless, you should know that by now.


Because, again, SR cannot distinguish between upside and downside volatility.

Here is an extreme example: If, on average, you make $1 MILLION on winning trades and lose $0.50 on losing trades (for instance), SR will say that your system is too risky and too volatile, a complete nonsense obviously.

Get it now?

Granted, some of them are good, but they are rare.

And in any case, don’t expect them to publish their money-making trading secrets/findings in a book, magazine or online, they are not that crazy…

you can do it this way if you want.
c2 by the way excludes risk-free return and calculate system return to drawdown ratio.

1 Like

The problem with sharpe or any ratio on most C2 systems is that it is based on a few years of data or less. If you look at the sharpe of S&P500 vs that of long term bonds there is a lot of history and significance.

Sharpe on C2 doesn’t mean much to me.

The best C2 metrics in my opinion are does the philosophy of the leader make sense, are they taking on too much risk in terms of position sizing, and if both of those have good responses then I would look at performance.

I think sharpe is very meaningful if it is based on lots and lots of data and a long time period etc. A few years isn’t enough for it to mean a lot to me.

Just my opinion.


Hi Dwight,

If the C2 trade leader system has no mathematical edge to begin with, no “position sizing” or “money management skill” will ever save him from a 100% guaranteed failure, period.

This is something that the vast majority of the so called “experienced” traders have a hard time understanding.

It’s like playing roulette in Vegas or Monte Carlo, no matter how “conservative” your system is. you WILL lose in the long run, because you have no edge to begin with.

I largely agree with you but at the same time equities over the last hundred years have a tendency to trend up and make you money. Casinos have an edge against you. I think casinos are great for analogies but the mathematics are very different in my opinion. Stocks and bonds are much more generous than the casino.

If a system could only either hold 100% SPY with no margin or 100% cash it would be very easy to underperform the market, but it would actually be very difficult to make a 90% drawdown in the account in a year. If someone can find a way to do that they would actually have a great short system.

There are positioning rules that make it very hard to destroy the account. Without a trading edge it is easy to underperform, but good position sizing can protect agains big catastrophic damage.

If instead of doing only 100% SPY the system could do 2,500% SPY it becomes very easy to have a 90% drawdown in days.

That is what I mean by position sizing.

To say it another way imagine that you are at a casino and have $100 to start and can play a game where the odds are in your favor 60% and against you 40% - great odds. If you double your money when you win and lose your bet when you lose you can still blow up your account with bad position sizing. You could bet $50 in a row twice and have about a 16% chance of blowing your account up completely - even though you have a 60 to 40 mathematical advantage.

You need good position sizing and an edge. Fortunately the markets already have an edge that helps investors.


True, over the long term, the market returns about 10% a year on average.
Please note that half of that comes from dividends alone, and if you add inflation and taxes it’s more like 4 ot 5 % a year.

In other words, it’s nothing to brag about, unfortunately, but it is still better than receiving 0% each year, if we keep our money under the mattress (LOL).

I couldn’t agree more on that one my friend, position sizing IS indeed the key to success, assuming the trading system has some predictive power in the first place.

By the way, I strongly encourage our C2 traders to read “THE SUCCESSFUL TRADER’S
GUIDE to MONEY MANAGEMENT” (Andrea Unger). This fantastic book explains the extreme importance of position sizing, and will surely amaze you.

Nice post overall Dwight, thank you. :v:

Thank you! I’ll have to check that out!

1 Like

I think you misunderstand what SR is. You only explain the standard deviation part and you don’t explain the compounded return part. In your example, despite the volatility is high, the SR should be larger than 2 because the strategy makes a huge compounded return.

Of course, most papers are useless, but I don’t t think retail traders know more than academic researchers. If you want proof, just randomly ask a retail trader about market microstructure and see what he knows.

On January 1st (2020) I have exactly $5,999,997 in my trading account. The risk-free rate in the US is 1% (for example).

Here are my trading results for 2020:

Month 1: -$0.50
Month 2: +$1,000,000
Month 3: -$0.50
Month 4: +$1,000,000
Month 5: -$0.50
Month 6: +$1,000,000
Month 7: -$0.50
Month 8: +$1,000,000
Month 9: -$0.50
Month 10: +$1,000,000
Month 11: -$0.50
Month 12: +$1,000,000

Total for 2020 : +$5,999,997 (a 100% return on capital)
The standard deviation of this system is 500000.25 and its Sharpe ratio is 0.

So according to the Sharpe ratio this is an extremely bad trading system, even though I doubled my money in 12 months and my losses are ridiculously small.

what are u talking about? :joy: :joy:

the return of that year is about 5.99 million while the standard deviation is just about 0.5 million. How can the SR be 0??? The SR should be close to 12