Estimating the risk of inflation in property investments Free

Estimating the risk of inflation in property investments

Purpose – Property returns are normally measured against the target rate for similar investments with comparable risks and liquidity. However, this analysis is normally undertaken in nominal terms and thus the risk of inflation, as it affects different investments, is not fully quantified. This paper seeks to analyse the effect of inflation on property investments.Design/methodology/approach – This article examines the impact of inflation on gilt returns and relates this to property risk.Findings – Investors may take a more pessimistic view of future inflation as an investment risk than the current official indices would indicate. In this context it may be that retail price index (RPI) and index adjusted for mortgage payments (RPIX) are not reliable indicators of inflation risk. It has been suggested that the difference between the two species of gilts as “a calculation of inflation expectations should be regarded with suspicion because of the volume of index linked bonds is so small that individual trades can move the market”.Practical implications – Economists and financial advisers and commentators have long recognised that inflation, in the sense of the tendency of the value of a currency to decline in purchasing power, distorts the picture of the worth, not only of individual assets but also of the whole economy. In this respect investment advisers often, in presenting their arguments, use yields that are net of the rate of experienced inflation taken from the performance of the RPI or the RPIX. Unless there is an understanding of the risk of inflation on property investments, such net rates may be misleading.Originality/value – This study adds to the literature exploring the effect of inflation on property returns.

Keywords: Inflation, Retail price index

Introduction

It is widely accepted that rational appraisal and valuation methods for property investment purposes are those based on discounted cash flow (DCF)techniques[1]. The essence of DCF is the discounting of expected cash flows,both of income and capital, over a future time or project life. The rate of discount is derived from comparison with an alternative “target yield”for investments having similar or comparable risks. It is further recommended that the basis of the comparative test is to adjust the yields used, by appropriate premiums, for the differences in risks. In particular in a real value approach these risks include the presence of expected or anticipated inflation and to use net of tax cash flows. These are the two most significant differences between business practice and property investment practice.

Measuring interest rates

The main issues in the comparative analysis between gilt edged investment investments and property investments are in the frequency, timing and nature of the cash flows and differences in the marketability and risks associated with the two species of investment. Some adherents of the conventional property investment approaches (Lumby, 1984; Carsberg, 1974) believe that property investment yields may be compared with those from the market for stocks and shares. The relationships between the various investment markets are not simple and straightforward. It is widely accepted that the basis of the comparative tests is to adjust the yields used, by appropriate premiums, for the differences in risks to security and variability in the cash flows. Differences in the timing and components of cash flows can be adjusted in the mathematical techniques used.

In the UK, along with most western economies, there are efficient and open markets for state bonds, i.e. gilts. These are investments that are relatively free from entrepreneurial risks (Fraser, 1977, 1986) and whose only risks may be associated with, or arise from one or more of:

• liquidity;

• costs of investing and realisation;

• taxation;

• inflation; and

• government financial failure.

With the exception of the last of these all these factors have some predictable range or scale of probability, which economic forecasters are able and willing to estimate. The same cannot be said of the risks in individual business ventures or property transactions. Hence the primary source of reference yields for major investors in the UK derive from the markets for gilts.

The problem of inflation risk

Economists and financial advisers and commentators have long recognised that inflation, in the sense of the tendency of the value of a currency to decline in purchasing power, distorts the picture, not only of the worth of individual assets but of the whole economy. In this respect investment advisers often, in presenting their arguments, use yields that are net of the rate of experienced inflation taken from the performance of the retail price index (RPI) or the index adjusted for mortgage payments (RPIX)[2]. It is now virtually standard practice for reports by financial commentators[3] comparing the aspects of the economy on a real value basis making ex post adjustments. For decision takers and their advisers the problem is one of the expectations of inflation to be determined ex ante[4]. The best bonds for these purposes are undated[5] fixed interest gilts (FIG), such as 2.5 per cent consolidated stock or 4 per cent treasury bonds. The advantage of FIG as a measure of a reference yield is that there is no potential for increase in value[6], other than to the capital by market trading, and their only relevant risk is inflation, which depletes their capital value and income. While FIG provided a good benchmark they do not, of themselves, have any means for assessing the risk of inflation to the investment. This changed when, in 1981, the treasury issued the first tranche of £1,000 million index linked gilts (ILG)[7]. Since then there have been 12 further issues[8]. Because they are index linked all the ILG are redemption stocks. Some commentators opined that the market for gilts would be able to provide a reliable measure of the expectation, i.e. risk, of inflation[9]. In the event the resolution of the problem has not proved as simple as expected. In order to answer criticisms of using an inflation risk free yield (IRFY) in property valuation and appraisal the author undertook this study of the relevant gilt yields in comparison with experienced inflation. This provided the opportunity to illustrate what others have suggested[10].

The analysis of inflation risk

The data eventually used for the study was obtained from published yields for the 12 months from February 1996 to January 1998. This meant that data had to be collected to provide the comparison. The first set of data covered the 12 months from February 1996 to January 1997. This data provided a comparison of the relative performance of the two kinds of gilts, i.e. ILG and FIG. The following four sets of stock selected on the basis of comparable redemption dates are shown in Table I.

The differences in the performance were compared with changes in the RPI as set out in Table II. These differences were greater than could be explained either by current rates of experienced inflation or from the range of inflation for the short term predicted by economic forecasters.

The difference is most striking in the case of short dated (only one or two years). By way of example, Figure 1 illustrates the difference for such bonds. Forecasts for the short-term, when inflation is at its most predictable, are the most reliable. The differences do, however, fall within the upper range of the Bank of England’s[11] forecast.

Explanations

Among the possible explanations for these differences are:

• Investors may take a more pessimistic view of future inflation as an investment risk than the current official indices would indicate. In this context it may be that RPI and RPIX are not reliable indicators of inflation risk.

• It has been suggested that the difference between the two species of gilts as “a calculation of inflation expectations should be regarded with suspicion because of the volume of index linked bonds is so small that individual trades can move the market”[12].

• The way in which the index linking of the interest payments and redemption price is accomplished[13] may be a factor in the difference.

Commentary on these explanations

Given that the dealings in gilts take place under near perfect market conditions and are carried out by sophisticated and informed investors this is not likely to be even a partial explanation.

With regard to this explanation, by an unidentified author but in a highly regarded publication, no evidence was given to support this thesis. It is not one that was indicated by authorities such as Professor McRae[14]. Every deal in a market place has the potential to affect prices in that market although such an effect may not be statistically significant. ILG represent about 13 per cent of all government bond issues, which is not an insignificant proportion. It may be that the deals are restricted to a small number of investors. If this means that the relative prices of ILG are consistently higher than those for FIG, with similar redemption dates, then there are two possibilities. Either the market for ILG has a different efficiency from that of the market for conventional FIG or there is some difference between the bonds, which is not solely, or directly due to investors’ expectations of inflation risk. This possibility could lead to a consideration of the motives and needs of funds that are likely to invest in ILG. This is considered later in this report under the heading of investment in indexed linked gilts, infra.

The last of our possible explanations could be associated with the linkage between inflation performance and the two species of gilts. This is considered further under inflation linkage between ILG and FIG, infra.

Investment in indexed linked gilts

It is possible that the principal investors in ILG are mainly pension funds and life insurance companies for whom such inflation proofed investments may offer a special attraction[15]. These institutions hold large and diversified investment portfolios and face strong competitive pressures to try to maintain the value of their portfolios against the needs of their pensioners and members. It has been government policy, in the public sector, pension payments have to meet the rate of inflation only and so for “funded” schemes the portfolios must maintain the value of the funds’ obligations in real terms. ILG would seem to be the ideal vehicle for this purpose. This advantage might be sufficient to make it worthwhile for Pension Funds to pay a competitive gilt market premium (GMP) over and above the inflation risk.

Estimating a gilt market premium

This notion of a GMP to investments in prime property, which are more narrowly traded in the far more inefficient and imperfect markets than those for gilts, could be a premium that might be aggregated with the property market premium. This latter premium is in respect of the marketing advantages of gilts over property investments. The quantum of this premium has been identified as being between 1.5 to 2 per cent[16]. This then gives a basis for calculating the aggregate premium that investors might pay in determining property yields to cover the risks in property over those in gilts. This would be by using the yields on ILG as the target yield for property comparisons, but adjusted by the aggregation of the two premiums.

This argument can be illustrated mathematically with the yield r_fon fixed interest stock used as a basis for calculating the equated yield e[17],can be presented in the following form for use in tables and calculations to take account of specific rent review patterns and value change.

Equated yield analysis and formulae

To calculate the value of the equated the following formula can be used:in which:In order to further analyse the role of a GMP[18], it was assumed that short dated ILG, of which there is only one example in this study, with a redemption date of 1998,carries the most predictable or accurate rate of inflation risk. If the risk of inflation to such stock were assumed to approximate to the RPI this would give the value of m as the difference in yields between the two species reduced by the rate of the RPI. Table II shows the mean value of the difference of the short dated stock is 4.3 per cent with SD of 0.3 per cent. This average yield on short dated FIG after reducing it by the rate of inflation gives a special market premium in the range of 1.3 to 2.6 per cent with a mean of 1.8 per cent. The GMP was then applied to each of the differences for the other stock to provide a measure of inflation risk for the medium, long and very long terms as given in Tables III-V, and in Figure 2[19].

The adjusted figures may be a more realistic expectation of inflation risk but there is no certainty that the GMP will always explain the difference as they may be due to factors that have not yet been identified.

Inflation linkage between ILG and FIG

Apart from assuming that the additional yield of FIG over ILG as an analogue of inflation risk there is another promising approach. This is exemplified in the table of figures of the “breakeven inflation rate” published in the monthly journal, Money Facts[20]. These are, effectively, of the rate of inflation that would be needed to equate ILG yields to FIG yields with a comparable redemption date. The publication of these figures obviates the need for calculations to deal with the indexation.

To examine this a further study of FIG and ILG yields against experienced inflation was carried out for the period from August 1997 to May 1998. This showed that the level of risk of inflation was about 3 per cent for both the medium dated (2001-2009) and long dated (2011-2030) stock (see Figure 3).

This estimate was also in line with market-based observations of inflation risk. The results are not significantly different from those based on the techniques of yield differences previously demonstrated (see Table VI).

There is, however, a slight drawback to the use of the approach for this purpose. This is that the measurement of the increase in yield is restricted to the medium and long dated gilts (i.e. excluding short dated and undated, i.e. perpetuity). The omission of the short dated is not a serious problem as short-term forecasts are regularly available in the financial press. It is difficult, however, given the nature of the statistics involved, to envisage a statistical trend line from the published figures. Instead they present a snapshot at the relevant date rather than a trend line taken over a series of dates. Having said that their publication certainly helps to corroborate other estimates.

Conclusions

There are at least four possible sources for estimating inflation risk. These are:

• 1.

  Published short-term forecasts by official sources, e.g. Bank of England[21].

• 2.

  Published forecasts by universities and financial observers based on economic modelling[22].

• 3.

  Estimates based on comparison of FIG and ILG as in the study illustrated above.

• 4.

  Estimates of “breakeven” inflation rates, such as those published in Money Facts.

When giving, solely, a commentary or explanation of past economic performance ex post inflation rates can be taken directly from published statistics. For those advising on, or having to take a view of, inflation risk any one or more of the above kinds of analyses have some validity and usefulness. This is not to avow that the analysed components of the risks of value change are hard facts, they are estimates based on explicit and rational hypotheses. Estimating inflation risk is not such a serious obstacle especially where the alternative techniques support one another. Overcoming the problem of estimating this particular risk should be no more daunting and insurmountable than coping with the other unknowns and difficulties in valuation and appraisal. These include all future time based variables. These include accurate information on market conditions, taxation, etc. There are also the usual problems of obtaining reliable data from relevant transactions of which only a few are in the public domain. All of this leaves the valuer having to take account of unreliable data on prices, rents, financing, business strategies, age and condition of buildings, use values, etc. This puts into perspective the difficulties of estimating inflation risk against the background of deficiencies and illogical procedures that may arise by relying exclusively on ARY methods,which cannot take account of value change.

The rates of future value change, i.e. whether based on inflation or on market factors, can never be known with certainty but the kind of analysis examined here demonstrates that expectations of inflation can be predicted within a range of probability. The other causes of value change will remain a matter of professional judgement. This means that it is possible to relate property investment yields to the yields on gilts. Once a rational decision has been made on the choice of reference yield and on the ranges of the risks and prospects of value change comparison between gilts and property yields.

The three DCF techniques are the internal rate of return (IRR), the net present value (NPV), and the annual equivalent (AE) (Merrett and Sykes, 1963). These authors refer to these “techniques”as “methods of investment appraisal” ibid. chapter 2.

E.g. As a basis for measuring the performance of a portfolio of shares (Macfie, 1994). See also Pike and Dobbins (1986) especially chapter 6 contributed by Professor McRae.

E.g. Gavyn Davies of Goldman Sachs.

Cf. Bank of England Quarterly Bulletin, December 1983 at p. 471, also footnote p. 38, ibid.

Dated stock yields may be used but the “flat” yield must be adjusted for redemption. Yields with such adjustments are normally given in the financial press.

All “increases” in value are often loosely referred to as “growth”. To avoid ambiguity it is better to distinguish between real value increases and those solely due to inflation.

Index linked retirement bonds were actually issued in April 1975.

See Rutterford (1983)for an explanation of the indexing mechanism.

E.g. McRae in Pike/Dobbins op.cit. “… the current value of these (index linked) bonds compared to similar non-inflation proofed bonds gives a good estimate of market expectations as to future average inflation … ”

Pike and Dobbins ibid.

The bank’s forecasts of inflation in February 1997 charted the range of RPIX for the ensuing year at the 25-50 per cent confidence level, at about 2.1 to 3.5 per cent but between about 1.5 and 4.6 per cent at the 90 per cent level.

Hamilton (1994). The same reference makes the comment that “implicitly (investment decisions are based on real interest rates(it is the expected real interest rate that influences behaviour”.

Rutterford, op.cit. gives a concise explanation of the rules for indexation.

Pike/Dobbins op.cit.

The Goode Report on pensions in 1995 recommended that funds should increase their holdings of fixed-interest gilts relative to equities and claimed that index-linked would suit mature funds.

Baum and Crosby (1988) and Fraser (1986). In September 1982 it was reported in the Estates Gazette (Vol. 263, p. 763) that a leading surveying practice estimated the difference between the yields from ILG and prime property at 2 per cent.

This was the reference yield recommended in Trott (1980).

Represented in the following formulae by the symbol m.

From statistical “linear trend lines” the indications are that, at the time of the study, in the short-term (up to three years) inflation risk is tending to rise but with the medium term (3 to 15 years) and the long-term (>15 years) the trend is downwards.

Provided by Barclays Capital.

The Bank of England forecasts are normally over the ensuing two or three years.

As Professor McRea, in Pike and Dobbins, asserts there is always the option to take the average of published predictions on inflation made by economic forecasters for the period in question.

Abbreviations: e; =equated yield.; r_f; =yield on fixed interest gilts.; r_pp; =rate of the property market premium.; r_I; =yield on index linked gilts.; r_m;=rate of the gilt market premium.; d; =rate of expected inflation.

Ernest WoodSouthport, UK