Back

"Expectancy" meaning on the Overview tab

Does anybody know what the "Expectancy" definition is (on Overview tab of the results)?
I did not find it in the documentation.
Update Backtest






The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.



Expectancy is used to estimate the likely outcome of a trades over time. If its positive the strategy will probably be profitable even if the loss rate is high.

[tex]Expectancy = Probability Win * Size Win - Probability Loss * Size Loss[/tex]

Link
0

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Agree - this is my understanding too, here is the result I have:
Average Win 0.7% Win Rate 72% (0.7*.72 = 0.54)
Average Loss-0.2% Loss Rate 28% (0.2*0.28 = 0.056)
Expectancy 2.195
So this means that 2.195 would be achieved if instead of "%" for average win or loss is used actual value in points.
Otherwise the expectancy (in %) should be: 0.54-0.056 = 0.484%

Since price of securities/ETFs vary a lot, I would say that the Expectancy in points would give little information, but expectancy in "%" would be better.

Expectancy in points would (in my opinion) make good sense when provided for for futures.
0

Yes I agree Nik it feels more sensible in percentages, but everywhere I see it explained is in dollar terms. Here is the line which calculates it - which I think is a slight refactor on the original formula. If you have a textbook/reference which shows it in percentages I'd be happy to change it.
0

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


I googled "trade expectancy in percent or point"and http://www.financial-spread-betting.com/Expectancy.html uses percentage in both multipliers.

I agree that most sites use dollars as you mentioned. But this is because their explanation always starts with something like: "Assume you have a system that wins 40% of time $1000 and loses 60% of times %500 ..." in order to make the point that you can have a winning system that actually loses most of the time.
Again for me $$ instead % makes more sense for futures, where you would trade e.g. 1 contract for every $10,000 you have in the account.
Not sure if others feel the same though ...
0

With clean in-out trading its easy, but for portfolios where you sell 10 shares on share 100,000 holdings the percentage doesn't have any meaning - e.g. percentage of holdings? It might be cleaner to just remove this statistic entirely as its much more trading focused and less quantitative.
0

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


In case of multiple buys and sells that can't be matched, perhaps we can use the average price of the buys and still calculate the expectancy when sell is made and then use it to get the average. I think for the majority of the developed algorithms it will work, since buys and sells match. For those that it does not work, the average buy price could be used.
And the same problem exists when you calculate avg winner/looser too (we would don't know the base price).

Since this thread turned slightly to direction of suggestions, i would add few more:
1. The STD is great but does not dive the visual if gains/loses over the tested period are well or badly distributed. I would suggest to add to results the portfolio gains for each tested year. This would immediately tell me the yearly gains deviation is good or bad in a visual way.
2. Another important one I would add is number of bars the trade lasted. I would do this for winners and loser separately. This will give feedback if I am "cutting my losers quickly and letting winners run", so it gives ways to improve strategy in the right direction.
0

Update Backtest





0

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Loading...

This discussion is closed