If you look at the books (1574489) or (1431387), they seem to leave a lot of cash on the table.
I have other books made from the same two strategies and they all leave a significant amount of cash on the table e.g. $7.2k on a $250k book which is significant (circa 3% to 3.5% of the value of the book)
Strategies are: 1574307 & 1431260
There no obvious reasons for this and I did not notice it in the past 4 years which lead me to think that this is new (usually would leave $50 cash or so).
Cost and slippage are set to 0 for both underlying strategies
Books are on variable slippage and 0.04% transaction costs. The T-costs seem to have come out all right today (see another post and email exchange with Ted)
The excess cash is due to the inability of books to buy fractional shares.
To simplify things, let’s say you had two strategies, one of which bought only AAPL ($195.59) and the other of which bought only GOOGL ($1,094.61). Let’s say each strategy started out with $50,000. Strategy A would leave $124.55 on the table and strategy B would leave $742.55 on the table. The total cash would be $867.10, or 0.87%. Now let’s say you started a book with $50,000 cash that combined those two strategies. Since the amount in the book is significantly less than the amount in the two simulations combined, you’d get significantly more cash (percentage-wise) in the book than in the two strategies. To be precise, you’d get $160.07 + $918.58 = $1,078.65, or 2.16%.
Thank you Yuval. I understand and the maths check out.
Indeed, if the book in your example was $200k, it would only be 0.22% left in cash (instead of 2.16% with a $50k book).
In general the larger the book, the less issues like this.
The annoying bit is that this process creates a gap from book simulations (which trade the equity curve of the underlying strategies) and, over time, it can lead to differing results… Not much can be done about it.
I assume fractional shares for books are not on the to-do list?