While I have no empirical evidence to quickly draw upon, it’s my belief that most firms either treat cash flows as “start-of-day” or “end-of-day” events. Such a convention is thought of as reasonable and even perhaps “best practice.”

Sadly, I must say that I believe it is wrong.

This post was prepared in response to an inquiry from our colleague, Irina Almeida, who had some questions about this subject. More will follow, as this is an important topic.

What’s wrong with start- or end-of-day cash flow timing?

If everyone’s doing it, why’s it wrong?

Before we look at the math, let’s consider the following: A client adds €1 million to their €5 million account. Some of this money is invested during the day, and the securities that are purchased with it increase in value by the end of the day.

If you use an end-of-day approach for your cash flow timing, then the just purchased securities’ appreciation will be credited to the €5 million starting value, resulting in an overstatement of the return.

What if those investments lost money during the day? Well, in this case, by crediting the loss to the starting value, you’ll penalize the manager, by overstating the drop in performance. Recall that the return formula is essentially gain divided by the beginning value (adjusted for flows). If you ignore the increase in working capital for the day, your denominator will be smaller than it should be, meaning you will inflate the gain or loss.

Let’s say that you’ve sold some securities during the day, in order to create cash that’s being transferred out. By treating this as a start-of-day event (that is, the transfer goes out at the beginning of the day), you’ll inflate the loss, since it’s against a smaller denominator.

What’s the solution for cash flow timing?

We recommend the following:

• Inflows: start-of-day

• Outflows: end-of-day.

We’ve been preaching it for more than ten years, and believe we’ve had a great deal of success at convincing firms, both software and asset managers, to adopt this approach. While this may seem a bit complicated, more and more firms have adopted this idea.

A cash flow timing example

This problem was sent to us recently by a client.

A portfolio begins the day with €8,087,171. An additional €87,752,553 comes in, and new purchases resulted that day. At the end of the day, the portfolio is valued at €95,554,640.00, meaning it suffered a loss of €285,084. What’s our return?

If we treat the flow as an end-of-day event, then the loss is applied only to the amount the portfolio began the day with, and the return is -3.53%; however, if we treat it as a start-of-day event, it’s -0.30%; i.e., a 323 basis point difference!

The loss occurred as a result of securities that were purchased with the new cash dropping in price by the end of the day. Clearly, this should be a start-of-day event; however, the client’s cash flow timing default was end-of-day for all flows.

We use similar examples in our Fundamentals of Performance Measurement course, and spend a bit of time discussing this subject. I think it’s fair to say that most individuals come away questioning their cash flow timing policy.

A common topic for discussion

The Performance Measurement  Forum meetings have spent time on this subject during several sessions. So many managers observe what are clearly erroneous results by using either the start-of-day approach, when outflows occur, and end-of-day, when inflows happen.

The problem becomes more noticeable when large flows occur, as with the example above, especially when accompanied by significant market movement.

Over the years, we’ve occasionally gotten calls from clients and others who question their results: in every case, it turned out they were using the wrong cash flow timing, and in every case, the start-of-day (inflows) and end-of-day (outflows) rule resulted in the correct returns being produced.

This might surprise you! Cash flow timing has only gotten more challenging with daily performance

Yes, it’s true!

As many firms have moved from monthly to daily performance, we’ve found that this problem becomes more prominent. It exists with monthly returns, too. And, with the GIPS® (Global Investment Performance Standards) requirement that firms revalue for large flows, it can also be felt more significantly, if the timing is wrong.

Have a different view? Please chime in!

Again, I’ll devote more time to this, most likely in our September newsletter, as well as in a forthcoming article.