Discrete vs cumulative performance data

Some fund managers argue that size is an issue and the impact can have potentially two issues: first, because of the quantity of shares a manager wants to purchase, they would most certainly move the price and second, they would have to ensure that they do not breach the takeover rules when investing even a relatively small percentage of a fund into a small company.

The takeover rules govern how much of a company can be owned by one investor without the requirement of a takeover bid.

Is big bad?

So do these issues necessarily mean big is bad? Most certainly not. There are several things to consider here:

First, what is the objective of the fund in question? If it is a large capitalisation, globally diversified fund, then a large fund is fine as it will have a large number of individual underlying holdings thereby diversifying security risk widely. So if one investment failed (and as you know, some do and we often cannot predict which one it will be), then having it as one of several thousand holdings will cause the failure to have a minimal impact on that portfolio’s performance.

Compare that to a fund with, say, 50 holdings, that follows the same investment strategy. The second fund cannot be as widely diversified and it is likely that the managers will have to take a view on which investments to include and exclude from their portfolio. If one of 50 holdings failed, then the fund would reduce by 1/50th, leaving alone how the other holdings might have performed. That only happens if the investments are equally weighted, and in most cases the exact percentage is likely to be bigger or smaller.

Some individuals believe that you can be over-diversified and that holding too many individual underlying holdings – perhaps more than 200 – does not diversify the fund any further; all it does is increase the costs of dealing and monitoring those holdings.

Again, it is back to what the fund is trying to achieve. If it is wide global diversification then this argument may not hold water. However, holding every listed security probably would be too many.

One of the advantages of larger funds is that they can approach a seller and offer to buy a large quantity of stock immediately, albeit at a reduced price – there is a trade-off between price, quantity and timing with any transaction. As a result they can sometimes achieve greater economies of scale than a smaller fund might be able to. Some funds often select their holdings based on certain characteristics, for example their sensitivity to particular risk factors, and do not necessarily review the company in minute detail by, say, interviewing the management before purchase.

To quote Buffett

Warren Buffett once said that the best investment was one you could hold forever. One of the sectors in which he regularly invested was insurance companies because they had a regular income stream. However, the problem with most investors is that they do not operate a long-term buy and hold strategy and most certainly do not hold forever.

Market research firm Dalbar carries out an annual study of the returns achieved by US investors and the 2014 results are shown in Graph 1. Interestingly, we can see that the average US investor did not receive anywhere near the return of the S&P 500 index over the period shown. A similar story is seen with the average bond investor relative to a bond index. Why is that?

It is a fact that most investors do not buy and hold their investments for the long term. So why does this happen? There are a number of reasons. The biggest impacts are poor market timing decisions and those trades being too numerous along with the associated costs of them. So what does that mean when we look at cumulative and discrete performance data?

Let’s take each in turn...

Cumulative data shows what an investor would have received if they had invested in that particular investment AND stayed in it for the entire period, whatever the period being displayed. For example this could be the average cumulative return over five years. Note that this data may or may not be annualised as some data providers show the cumulative returns over the entire period, five years say, so it might show +125 per cent which is the total return over the entire five years or 17.6 per cent, the annualised equivalent.

Discrete data meanwhile shows the return in each individual period separately, typically each individual year’s performance although discrete months and quarters are also common. Over a five-year period, there might be five individual performance figures, one for each of those years.

How are these figures useful? The answer depends on what you are looking for. Say for example you are looking for funds with a steady and consistent performance above a certain percentage. You might first look at cumulative performance to see which funds had achieved that return as an average annualised performance. However, what that does not show is the extent to which that performance was consistent over the period, in each and every one of those years. It might have trundled along the bottom of the performance charts for the majority of those years and had only one stellar month which then had the effect of boosting the average annualised returns (see fund A in Graph 2). If you were looking for a consistent steadily performing fund, this would not be it.

To avoid choosing incorrectly, we could turn to the discrete performance data because this would show what had happened in each individual year. This data would then reveal which funds had indeed rumbled along the bottom of performance tables and then shot up once, shooting the lights out.

Note here that the overall performance of both funds over the five-year period is the same; the difference is how each fund has got there. Other data such as the fund’s volatility (measured by its standard deviation) might give an indicator of this; remember what volatility is showing. Volatility is the average of a fund’s performance plotted against its mean.