Calculating compounded returns

[Kim 7,512]

One of SO former members posted a review on Investimonials - http://investimonials.com/newsletters/reviews-steadyoptions.aspx#alock

 

His main claim was inconsistency of the compounded results:

 

"Kim uses 10% of the balance on the following trade, with a starting amount of 10,000.

Take a look at the trade that was entered on 6/2/14. The balance is 16,953. This means he will use 1695 on the next trade. His next listed trade was entered on 5/29/14 and earned 25%. That is 423.75 off of the 1695. Add that to the balance (16,953) and you get 17,376. Looks correct but wait, the 5/29/14 trade used a balance that was from the future! Only after the 6/2/14 trade was closed on 6/9/14 did he have the 16,953 balance. How could he know what portfolio balance he would have 4 days in the future?

Look at the very first trades at the bottom of his table. On 12/30/13 he starts with 10% of the 10,000 portfolio, or 1000. He earns 21.4%, adding 214 to make his portfolio balance 10,214. On 12/24/13 he trades with 1021 and earns 21.8% but again, the 10,214 balance is 6 days in the future!

You can see this throughout his entire table."

 

My response was:

 

"The "big inconsistency" that you discovered is simply due to the fact that there are few trades open at any given time, and the order of the trades is by closing date. What you "forgot" to mention is the fact that it works both ways: in case of losing trade, the next trade is based on lower balance. For example, if the first trade has lost 20% instead of winning 20%, the next trade would be based on 980 instead of 1020. Since most trades are very short term, it should not make significant difference. In addition, when the trade is open, it can show higher unrealized gain than the final gain. But since we don't know the actual gain of open trades when we open the next trade, this method uses closed P/L of previous trades."

 

I realize that no matter how transparent and accurate our performance reporting is, there will always be someone trying to find "inaccuracies". However, I would like to get members feedback and suggestions how to fix this issue. Personally, I think that the only way to resolve it is not to have few trades open at the same time - this way the next trade will be always open after the previous one is closed and the new balance is known. While I believe the way it is calculated now has no material impact on the bottom line, I would still prefer to find a solution if possible.

 

Any suggestions are welcome. You are also welcome to comment on the review and/or post your own review.

[Yobryan 2]

The benefit to your members of working more than one trade at a time outweighs the difficultly of accurately reporting compound results.

But to make the calculations more conservative you could only compound the trades monthly or some other method. I would guess that many member use this method on their own. They either use a fixed portfolio size for several months or adjust monthly. I doubt anyone is calculating the compounded amount from trade to trade.

Regardless of this estimation which I wouldn't even call an inaccruracy, the numbers are transparent and strong.

[LeSolist 2]

My two cents here.... Options portfolio's performance is always tricky. I'm a member of other big options and stocks services (highly reputable services) and they have similar issues because there are always ongoing transactions (trades, dividends, reinvestment or not, etc) in the portfolios.  

 

To add even more transparency,  I'd simply list a break down each trades to provide the in and out prices of each contract and type (Put/Call).Then, it's up to the members to calculate the performance (ROIC, Portfolio on the time basis they want to) as they wish. That said, I know that we can always dig the info by going through all the posts one by one but it's time consuming. 

 

Another thing you might want to do is to simply specify the value of the portfolio at the beginning of the year and at the end  (Typical portfolio ex: $10k Jan 1st, 2013 and  $18,970 on Dec 31st, 2013 if I read your performance table correctly). I care for ROIC but I care even more for what's left in my pockets at the end of each year.

 

Finally, you may want to explain how all your returns are calculated (ROIC, SteadyOptions Monthly Returns, total etc) in a sub page as I'm sure that not all members are 100% comfortable with all the terminology, starting with me.

 

Hopefully this "investimonial review" won't ruin your weekend.

 

Cheers!

Edited July 4, 2014 by LeSolist

[Kim 7,512]

My original thought behind reporting portfolio value was to show what was possible with 10% allocation per trade and compounding. It will never be 100% accurate for a simple reason that you can never allocate exactly 10% per trade, especially for smaller portfolios. Someone with 10k portfolio cannot buy 2.5 contracts when buying a $4 spread.

 

That said, I like the idea of monthly compounding. I just did the calculations, and monthly compounding will reduce the 2014 compounded portfolio value from 17,486 to 16,891. I take members feedback very seriously, and to prevent any more criticism, I'm going to redo the whole performance page with the new method.

 

Members can refer to Performance Dissected topic for more details how we calculate performance.

[LeSolist 2]

Thanks for the "dissection" page... I read it but I cannot figure out differences that I get between the yearly returns of the performance matrix versus the trades list's returns. For 2013, no compounding, starting portfolio value of $10k, invested capital of $1.5k per trade, I get 82.4% vs 89.7% on your matrix. That won't change my opinion in regards to your outstanding performances but I'd like to understand so this post. Any idea? 

 

PS: The devil is in the details...

-1: there are also small differences on the monthly values  (less than 1%) between the matrix of your website and the "Pro-Trading-Profits" matrix. Again, no big deal but I like to understand numbers.

-2: I noticed small differences as well (less than 1%) on the trades list's returns vs the reported return (what you report upon closing trades).

[Kim 7,512]

The matrix is taken directly from Pro-Trading-Profits. The slight differences might be explained by several factors:

 

PTP calculation is based on the number of contracts which is the closest to 15% allocation, rounded down. 82.4% is based on exactly 15%. Also, they include the subscription fee. I'm pretty sure that when the subscription fee went up in 2014 from $99 to $125, they updated the returns retroactively, but I did not. $26 difference on 20k portfolio (their default) is 0.13% which can explain the differences in monthly returns.

 

Could you please give some examples in individual trades returns? 

[LeSolist 2]

Here are some examples probably coming from a chain of rounding errors

1-

2/10/12 01/17/13 40 Days RUT butterfly -56.8% 9,127

vs.

Total P/L: 56.1% loss.

 

2-

12/13/12 01/04/13 44 Days GLD calendar -7.3%

vs.

Final P/L: 7.1% loss (0.42/5.87) 

Edited July 7, 2014 by LeSolist

[Kim 7,512]

I believe both are typos. But the results in the table are actually worse than the real results, so you can see I was not trying to skew the results in my favor 

 

Hopefully there are not too many mistakes, but let me know if you find out more.

[Kim 7,512]

An interesting development:

 

The same user who posted the above described review posted now a comment on another review http://investimonials.com/newsletters/reviews-steadyoptions.aspx#mattmcp:

 

"Please read my review about how Kim publishes inaccurate returns on his site and is completely fine with it."

 

There is just one small problem: the original review was posted by user named Alock, while the comment was posted by user named al0615. This is obviously the same user - he just did not noticed that he was logged in under different user when posting a comment. And look at al0615's profile - he posted a glowing review for one of my competitors just few days earlier. 

 

I would really appreciate if members could post a comment on his review or post their own review, to show your support.