The first day of back-testing a system showed a $20.24 profit on $6,000 of initial capital, for a net profit percent of .34%, and an annual return % of 241.85%. I’m having a difficult time reconciling the two. For example, if I grow a $6,000 investment at .34% compounded daily for 250 days (about a full trading year), I end up with roughly $8,300 interest. So I’m confused as to what the 241.85% is about. Thanks very much.
You can use an Excel spreadsheet (or maybe some web site). The return results you mention seem to be using all calendar days, i.e. compound growth of 0.34% for 364 more days. The initial 6,000 grows to 20,570 which equates to an annual cagr of 242%.
I can't attach a spreadsheet to this post but here is a screen capture of what the results are (the actual calculations in the Excel cells I can send if you need)
P.S. even at the 250 day mark how did you do your calculations?? This simple spreadsheet shows a different number at days 250 or 251 than your $8,300.
@portfoliobuilder is right.
Also reading the manual is not bad idea at all, see “Note: Calculation method used for annual percentage returns”.
http://www.amibroker.com/guide/w_report.html
(scroll down to the bottom where you will find the answer).
Thank you so much for your time on this - really appreciated. The difference I think between your $13,970 and my “roughly” $8,300 is primarily due to my subtracting out the original $6,000 (to focus just on profit, and is comparable to what is used for the .34% factor), and to a tiny degree my actually using (by mistake) a .35 growth rather than the correct .34 which you used. But by far the biggest conceptual difference was my thinking that compounding for 250 days (roughly one full trading/calendar year) - rather than 365 days - was the most appropriate time frame. I’m really new at this, but I believe that use of 250 days of compounding and focusing only on the profit component would make most sense in discussing a trading system against e.g. other types of investments. But in any case, again, thanks for clarifying this for me.
Using 250 days is wrong because:
- interest is earned every calendar day (including weekends)
- price gaps happen much more often over weekends than between regular week days so friday-monday gap is statistically different than monday-tuesday gap
Thank you Tomasz.
Regards,
Robert (Bob) Lerner