Flashcards in portfolio risk and return Deck (14)
Assessing fund performance
Portfolio risk and return analysis involves plotting the overall time-weighted return from a portfolio over a period against the “riskiness” of the portfolio. The portfolio is then compared to its peer group and to simulated portfolios such as median trackers.
The diagram below shows the results of plotting average annual investment return achieved over recent periods against the corresponding risk measured in terms of standard deviations of returns for a group of 30 portfolios. Ideally we would like to achieve a higher return and a lower risk than the alternative portfolios. Here Portfolios A, B and C appear to have produced the best results in terms of high return and low risk.
The riskiness can be measured by the standard deviation of returns over historic sub-periods. One method is to consider the volatility implied by derivative contracts on the markets and individual stocks in which the portfolio is invested.
The standard deviation of monthly, quarterly or annual returns is often used in
practice. The use of standard deviations is consistent with the capital asset pricing model, but other measures or interpretations of risk could equally be used. In practice a combination of several different measures might be used to assess risk.
The aim is to help assess the investment manager by ascertaining if superior performance has been obtained by taking more risk or by superior market and stock selection/timing, and conversely for inferior performance.
Investment performance is a function of both return and risk, so a true and fair assessment of investment performance needs to allow for both factors. One way of doing this is to use risk-adjusted performance measures, examples of which were described in the previous chapter. These aim to quantify the under- or out-performance after removing the influence of either total risk as measured by the standard deviation, or systematic risk as measured by beta.
Hirers of investment managers would not
want excessive risk to be taken. However this is difficult to enforce in real time. In practice, control over risk may be achieved by guidelines on the asset allocation and stock weightings, and perhaps by specifying a maximum under-performance over specified periods.
Thus, the investment managers might be told that the benchmark equity allocation over the next year is 50% of the total fund, but that they are permitted to deviate from this benchmark by up to 10% of the total fund size in either direction. Alternatively, the limits for departures from the benchmark might be expressed as a percentage of the benchmark allocation to the particular asset class, which has the advantage of permitting a smaller absolute variation in the lower weighted asset classes.
Limits expressed in this way are sometimes referred to as load differences and load ratios. We will discuss them further in the next chapter.
Alternatively, the investment manager may be instructed that the investment return over any particular period cannot:
be less than a specified absolute figure
fall below that of the average of the peer group by more than a specified
Downside measures of risk such as the downward semi-standard deviation mentioned in Chapter 21 might be of use here. The expected shortfall and shortfall probability measures described in Subject A205 might also be used – the latter indicating the frequency with which under-performance occurs.
Implementing risk controls
If the results of a risk and return analysis are discussed with investment managers at six monthly intervals, then there will be some measure of control, as it is reasonable to assume that, if only because of dealing costs, the portfolio will not be radically changed over such a period.
The dealing and other costs associated with a large change in asset allocation are discussed in detail in Chapter 22 later in this course.
Some problems with relying on this approach are:
Creeping change in portfolio composition. A portfolio may change every six months by an amount which is not of concern each time but which is significant in total. More risky stocks may be included which do not immediately show up in the standard deviation of returns.
This is because the standard deviation will be calculated as an average over a period of time longer than the last six months. Hence it will take some time for any creeping change in the overall volatility of the portfolio to show up in the data.
￼ A successful investment manager may be treated unfairly by the measurement system. For example, the manager may be skilled in selecting the good stocks amongst those which the market views as risky, as measured by the implied volatility in the prices of derivative contracts. Or he may be skilled at selecting stocks and markets that will be volatile but will grow over the period.
Hence, just because a manager has chosen a high-risk portfolio that has recently under-performed does not mean that he is performing badly. If the ultimate return turns out to be better than that required on such a risky portfolio, then he has done well. A sensible measurement system therefore needs to reflect the skills of the manager it is being used to assess.
The manager may disagree with the market view of a stock’s or market’s prospects and/or the uncertainty attached to those prospects. Market prices are a result of a pricing process which is likely to result in 50% of investors viewing a price as too low, and 50% viewing it as too high. Perhaps only very few professional investors will see a particular market price as roughly fair.
Likewise, the implied volatility of a share is set by the market, which may not be 100% rational. Thus, even if the manager’s portfolio has a high implied volatility, it does not necessarily mean that it will actually experience high volatility in the future.
It is hard to determine which of the above applies to an investment manager. Ultimately it is important that the manager can demonstrate a realistic and convincing rationale for investment decisions.
So asking the investment manager what she has done and why is as important as quantifying exactly what the consequences have been in terms of investment performance. The investment performance over a short period of time might have been disappointing even though the manager pursued an entirely sensible investment approach. Equally, good performance could be due to luck rather than sensible investment decisions.
It is likewise sensible discuss the investment manager’s ongoing approach at the start of each investment period.
Valuation and performance of equity investments 2.1 Using market price
The most obvious way to measure the performance of an individual equity investment is via changes in its market price.
As the market price is the yardstick by which an equity investment would usually be regarded as a success or failure, it is hard to argue that it has real limitations. However, there are some refinements which should be considered:
It is the total return that is of interest to an investor and so dividend income must be allowed for.
It may be that an equity price is influenced by short-term concerns whereas many investors are more focussed on the long-term. However on the other hand the long term is clearly uncertain. An equity price represents the market’s best estimate of the value of future cashflows to the average investor. If this estimate can move significantly over the short term, it is because some item of news has changed the market’s estimate of long-term cashflows or from movements in the market price of risk.
Best estimates of the value of long-term cash flows do change over the short- term because of the uncertainty attached to long-term estimates and the flow of potentially significant news items. The only alternative path for share prices would be long periods of stability punctuated by discontinuous jumps. Thinly traded shares tend to behave more like this than shares that are more liquid.
A separate point is whether decreases or increases in share prices should, in themselves, be part of the control process. This is basically an investment management call rather than a matter for automatic reaction. Typically markets go through periods of several years in which underperformers and outperformers frequently reverse their relative performance, and then go through other periods when over- or under-performance can be very persistent. The first sort of market tends to favour “value” investors whereas the second favours “growth” investors.
Absolute net present value Using net present value
The net present value of an investment depends on many assumptions, which can result in a wide variety of results. The absolute result from an estimate of net present value may therefore be of limited use.
Different assumptions regarding future cashflows and the risk discount rate will lead to different estimates of the net present value at any point in time. In addition the corresponding changes in net present value estimates through time will be different. The estimated net present value of an asset will also typically differ from the market price where there is one. This is because the assumptions underlying the net present value estimate (eg with regard to future earnings growth) will generally differ from those implicit within the market price.
It is important to investigate where the assumptions differ from those implied by the market price, since the market price represents a best estimate of present value. It is, in a very approximate sense, an “average” of investors’ estimates of the net present value of the asset, reflecting the average of investors’ views on future cashflows, investors’ risk tolerances, investors’ tax positions, etc.
However a particular investor may have different risk, or perhaps tax, considerations to those which drive the market price. Differences between the market price and an investor’s estimate of net present value can derive from:
1. differences between a particular investor and the average investor, eg with regard to their attitude to risk or their tax position
2. other differences in assumptions. Different investors will always have differing expectations with regard to the future cashflows (profits, dividends, rents, etc) yielded by a particular asset. Thus, two otherwise identical investors may come up with different estimates of the value of an asset. These are part of the range of value estimates that give rise to the current market price.
It would be justifiable to allow for 1, if the differences could be identified reliably. Allowing for 2 involves the investor in active investment management.
Changes in net present value over time
The trend of net present value estimates, and in particular their relationship to market prices, may be helpful. For example if large differences between net present value estimates and market prices tended to diminish over time, then the model used to estimate the net present values would be a helpful way of assessing cheapness/dearness of stocks or markets.
If the difference diminishes over time, this suggests that the net present value estimate has been proved to be “correct”. We might therefore assume that any existing differences will likewise diminish in the future and so buy shares for which the market value is currently less than the net present value.
Using net asset value
The net asset value of a company, or the net asset value per share, is only one component of overall value. Typically, the majority of the value of a company arises from the future profits that it is expected to generate.
So, if other things are equal, a share with a higher proportion of its share price represented by net asset value should be cheaper than a share that has less asset backing.
However, other things are unlikely to be equal as the market will, in both cases, be attaching a full value to all future cashflow, including that resulting from holding all existing assets and liabilities.
The net asset value is an accounting number, so it is important to understand how it has arisen and to make appropriate adjustments.
For example a company that has expanded by acquisition will have acquired goodwill on its balance sheet, which will form part of its net asset value. A similar company that has only grown organically will appear to have a lower net asset value per share. Generally, goodwill will have to be removed in order to make valid comparisons.
The goodwill referred to here represents the excess of the value paid for a subsidiary company over the value to the acquiring company of the share of the assets purchased.
Some businesses require more assets than do others. For example a manufacturing business will generally require plant, premises and stock whereas a service business will typically require less assets. In the latter case, there will consequently be more “human capital”, which does not appear on the balance sheet. In addition, some companies will lease assets, whereas others may own them.
Comparisons of net asset value between companies in different sectors may inform predominantly about the difference between the sectors. Valid comparisons can therefore generally only be made between companies in the same sector.
Net assets surplus to those required to run the business may not attract full value. It is usually regarded as inefficient for a company to hold surplus assets, and also it is harder to maintain management discipline when there is a substantial asset cushion.
The role of management is to maximise the earnings from the available assets (on behalf of the shareholders), ie to “work the assets hard”. If the company has surplus plant, machinery and stocks then it is easy to cope with variations in turnover and order levels. Thus, they avoid one of the important management tasks of efficiently scheduling the use of available assets.
In this context, you may recall the discussion of stock/inventory policy within the short- term financial planning section of Chapter 6.
Risk-adjusted return on capital
the assessment of the performance of a company, rather than an individual equity or an equity portfolio.
There are many different ways of assessing the performance of a company and in practice a combination of approaches is used, reflecting the motivation for assessing the performance. The obvious way to assess performance is perhaps via the profits reported in the income statement (profit and loss account). However, this doesn’t allow explicitly for the capital invested in the business to generate those profits. We might therefore look at measures such as economic value added (net operating profit after tax less cost of capital). However, this doesn’t allow for the degree of risk involved in the generation of added value. Consequently, we might therefore decide to estimate some interpretation of risk-adjusted return on capital.
The CAPM would suggest that the risk-adjusted return on capital should be equal to the risk-free rate. If capital assets are priced correctly then returns in excess of the risk-free rate will only be generated from taking risk. The risk- adjusted return is the actual return, reduced by b (rmarket - rrisk -free ).
This idea is based on the CAPM security market line, which predicts the returns on different assets when asset markets are in equilibrium. Here r is the actual return
on the market portfolio, which is often proxied by a market index. Thus, if we are able to identify both the capital invested in a company and the return generated using that capital, we can in principle calculate the company’s return on capital. This could then be adjusted as described above to obtain an estimate of the company’s risk-adjusted return on capital, which can then be compared with the risk-free rate.
Generally a high actual projected return on capital using normal accounting measures and a starting measure of the market capitalisation, would imply, in the framework of the CAPM,
the successful creation of intangible assets and shareholder value. The company would therefore be an attractive investment proposition. Similarly, high historic figures imply historic creation of goodwill and shareholder value.
Here intangible assets and goodwill both refer to any non-monetary and non-physical resources that are controlled by the company and from which future economic benefits (ie cashflows and asset growth) are expected. For example, successfully marketing a new brand and obtaining new customers both create value, but neither may appear explicitly on the balance sheet.
Clearly the appropriate b is an area where there can be different views.
Although historical values of beta can always be estimated from past price data, these estimates will be subject to the usual problems of statistical inference – eg choice of time period, random variation, etc. When assessing projected returns, however, we need forward-looking estimates of beta, which must always be subjective.
Calculation of capital and return
However it is difficult in practice to identify the appropriate return and capital figures for a rigorous CAPM-type return on capital calculation.
For example rather than excluding goodwill from all companies and transactions, as we suggested in Section 2.3 above, we have to include it in the capital for all companies, including in circumstances where normal accounting would not recognise a goodwill item.
This is usually the case with internally generated goodwill, which consists of intangible assets that are “self-created” – ie they are not purchased, for example via a merger or takeover. Possible examples of internally generated goodwill include customer relationships and brand names.
Consider the successful acquisition of a new customer or client, in a manufacturing or service business. This would be a balance sheet event for a CAPM calculation as an internally generated goodwill asset would have been created. However, the increment to value arising on this event would not be part of the return element of the CAPM calculation. The value attributed to such assets is just the number which, given the future incremental cashflow from the new customer, makes the risk-adjusted return on capital going forward equal to the risk-free rate.
It can therefore be very difficult to estimate the value of “capital” when attempting to calculate the risk-adjusted return on capital.
Ultimately, the best measure of the capital in a company, including all intangibles as required by the CAPM, probably is the market capitalisation itself. Building this up from historical transactions, including all intangibles is, though, a complex, and perhaps unnecessary, process.
Return can be calculated as profits . capital
The return part of the calculation also has to be adjusted from the starting point of accounting profits.
For example items such as advertising expenditure, which may all be expensed in accounts, will in part have to be treated as investment expenditure. The part aimed at expansion, will have to be added back to profits to calculate the return, or numerator, in the return on capital calculation. Only the expenditure necessary to defend existing goodwill should be charged to profits. Similar comments apply to investment in tangible assets.
Thus there is also much scope for judgement in deciding exactly what constitutes “return” when attempting to calculate the risk-adjusted return on capital.
According to the CAPM, return in excess of the risk free rate will only arise through taking risk. This does not mean that management cannot add value. Management’s job may be seen as the creation of internally generated goodwill in excess of the investment on intangible assets.
Management also needs to defend existing goodwill. An important part of all these efforts is well-directed investment, on both tangible and intangible assets.
Sometimes internally generated goodwill will automatically be recognised in accounts. For example if a property company lets an hitherto vacant building then the increment to value will be recognised next time the building is revalued. For a listed property company this may, depending on the listing authority, be every year.