They have a commingled vehicle that restricts cash flows to always occur on the first of the month. For contributions, this works quite well. However, they've discovered an issue with withdrawals.
One of their clients requested a sizable withdrawal last month. The money flowed out on July 1, which is fine and dandy. BUT, the money had to be raised during the month of June, meaning that for much of the month, their cash position was much higher than it should have been (it is typically close to zero, but was more than 10% as a result of the sale of assets to raise the requested cash). Consequently, the fund's return was much lower than it should have been, because of the cash drag.
A possible solution: once the cash was raised, it is no longer under the manager's discretion (i.e., they cannot invest it). And so, perhaps they should either (a) move the cash to a "temporary account" or (b) flag it as "un-managed."
Here's an example; hopefully it will help. It's in three parts.
Part #1: the portfolio as it would have looked at the beginning and end of the month, had there been no cash flow:
As you can see, we began with 100 million and ended with 105 million: a return of 5.00 percent.
Part #2: One client requested that 10 million be raised, and so, securities were sold mid-way through the month, in a proportionate manner, to raise the necessary cash, which remained in the portfolio until the end of the month [NOTE: if you have a hard time reading this graphic, just click on it to expand it]:
We see that the 10 million was raised halfway through the month and that it did not appreciate any further (we're assuming a 0% return for simplicity). Because the cash remained, its presence dragged down the return of the portfolio
Part #3: Finally, we see the case where we transfer the cash that was raised out (either to a temporary portfolio or to an un-managed status):
As you can see, by shifting the cash out, we are able to eliminate the impact it would have on the fund's return.
There are clearly some assumptions made here. One being that the manager has the ability to segregate the cash (which most accounting systems should be able to support), and provides returns to the fund's shareholders. Second, that the custodian's report won't cause any issues; it shouldn't, as the manager can explain the math behind his/her returns. And third, that the assets are liquid enough that mid-month valuations can be done, to support the bifurcating of the monthly return (prior to and after the transfer of cash).
If a firm does this, then it should document it and ensure that this approach is used consistently. It would seem to be applicable to hedge funds that might have to raise cash that they normally wouldn't have, and don't wish the cash to cause a drag on performance.
Note that the monthly return that is derived from the unit values (produced by the custodian) won't match what the manger provides; however, once we get past this month and calculate returns, the cash's impact should disappear. The only time it should have a lasting impression will be if this is done in December, in which case the manager's return will be the correct one; at least in my view. What's yours?
p.s., You may noticed that more than $10 million left; I show the full cash going out. We could have had just the $10 million go, leaving a small residual amount, but this would have been in the earlier case, too; the results won't really differ. I think the point is clear.