UK asset class returns

The London Share Price Database (LSPD) maintained at London Business School has served to form the basis for indexes that track UK asset class returns for three basic assets: UK stocks, UK bonds, and UK treasury bills from the year 1900 to the present.

Asset class returns since 1900


The long term returns produced by UK markets from 1900 to 2011 (as measured by ABN AMRO/LBS indexes) have rewarded investors with positive long term compounded returns. Figure 1. shows the cumulative increase in stock, bond, and bill returns. Over the past 111 years, UK equities provided a nominal compound return of 9.4%; long-term gilts provided a nominal compound return of 5.5%; and treasury bills produced a nominal compound return of 5.0%. These returns were consistent with the volatility of each investment. As measured by standard deviation, higher returns were correlated to higher risk.


 * UK stock standard deviation : 19.9%
 * UK bond standard deviation : 13.8%
 * UK treasury bill standard deviation: 6.4%

Of more importance to investors is the after inflation real return of investment. As figure 2 shows, the real returns of UK asset classes over the past century were 5.2% for UK equities; 1.5% for UK bonds, and 1.0% for UK bills. (These returns do not reflect the costs of investing nor the tax due on investment income).

These very long term compound return rates are the result of considerable shorter term variance. Over twenty year periods, the real return of UK stock market has ranged from -1.7% real compound return to 13.5% real compound return. The table below shows the real returns provided by UK markets over each decade of the twentieth century.

Downside risk
The downside risk of UK stocks and bonds can be seen in the historical drawdown experience from market peak to market nadir, and ensuing recovery of value (See Figure 3. 1930-2010). The real drawdown experienced by the UK conventional gilts market reflects the eroding effect of inflation on nominal bond purchasing power.


 * Stock market drawdowns (1930-2010)

The UK stock market hit a peak in 1936, from which stocks fell by a real -59% by June 1940. Stocks recovered value in 1945.

The UK stock market fell a real -74% from August 1972 to January 1975. Stocks recovered value in 1983.

The UK stock market fell a real -49% in March 2003. The recovery in value came in October 2006.

The UK stock market fell a real -47% in June 2007.


 * Bond market drawdowns (1930-2010)

The UK conventional gilt bond market fell a real -33% from January 1935 to September 1939. Bonds recovered lost value in April 1949.

The UK conventional gilt bond market fell a real -73% from October 1946 to December 1974. Bonds recovered lost value in December 1993.

Drawdown risk is moderated by diversifying investment between asset classes. A balanced 50% UK stocks/ 50% UK bonds portfolio has historically provided a standard deviation of 14.4%, virtually never suffered a decline of over 50%, and has experienced quicker recoveries of lost value when drawdowns occur.

Asset class returns from 1955-2000
Using share price and listing information in the London Share Price Database (LSPD) maintained at London Business School, a new set of indices for UK financial asset returns, starting in 1955, was created. These indices allow researchers to examine the cross section of stock market returns, as well as examine differing maturities of bonds.

The indices cover equities, including large, small, and micro-cap stocks. Examination of the value and small cap premiums in UK stock market returns have also been measured. In addition to a treasury bill index, the bond market indices offer separate coverage of intermediate conventional gilt, long term  conventional gilt, and index-linked gilt bonds. An index of inflation was also calculated. The data are consistent with the CRSP/Ibbotson series for the US.

Stocks
UK market composition over the 1955-1999 period, as defined by the ABN AMRO/LBS Equity Index, is shown in the table below. Large cap stocks comprise 90% of the UK market in terms of market capitalisation, but represent only 1.50% of UK stocks. Small cap stocks, comprising the next 9% of market capitalisation, represent 5.43% of UK stocks. While micro cap stocks make up only 1% of the UK market by market cap, they represent 92.97% of UK companies.

The following two tables provide summary returns for UK equity asset classes from 1955-1999. The first table provides nominal returns; the second table provides real returns. Notable observations regarding this data include:


 * The importance of dividends to total returns can be seen in the capital return component of the return in both nominal and real total returns. The difference between the two figures represents the dividend, and reinvestment of dividend, component of the return.
 * The historical returns show a premium return for small and microcap stocks. There has also been a value premium (0.49% monthly mean arithmetic return) for both large cap and small cap UK stocks. After the small cap premium was discovered in the 1980's, UK small cap stocks underperformed large cap stocks through the end of the twentieth century. The UK value premium was remarkably steady from the 1950's through the 1980's. Subsequent to discovery, the value premium has persisted, although it has become more volatile. Capturing the small and small value premium in UK stocks is problematic due to the higher transaction costs and greater illiquidity of the UK small cap marketplace.
 * Real annual returns in the UK equity markets have ranged from -60% to + 97%.

Bonds
The ABN/AMRO/LBS indexes provide for an examination of treasury bill, intermediate term conventional gilt, and long term conventional gilt total investment returns, as well as inflation rates. The indexes are comprised of the following fixed income securities.

The tables below provide summary returns for treasury bills, intermediate term gilts, long term gilts, and the compound inflation rate for the period 1955-1999. The first table provides nominal returns; the second table supplies after-inflation real returns.


 * The nominal return provided by bills and conventional bonds (8.3% to 8.9%) was greatly reduced by the 6.3% inflation rate over the last half of the twentieth century. The resulting real returns (1.8% to 2.4%) do not reflect the costs of investing nor tax on investment return.
 * Over the measurement period intermediate term conventional gilts provided higher returns (2.4%) than long term conventional gilts (2.1%) with less volatility. Long term gilts annual real returns ranged between -31.0% and 45.0%. Intermediate term conventional gilts annual real returns ranged between -12.0% to 29.0%.

Inflation indexed bonds
Index linked bonds first became freely available in the UK in 1981. An ABN/AMRO/LBS index of inflation-linked gilts has been calculated based on the following securities.

The table below provides nominal and real returns for inflation-indexed gilts.


 * From inception to 2000, inflation-indexed gilts have produced the lowest real return among UK government securities. From 1981 to 2000, long term conventional gilts provided a real 8.4% annual return; intermediate conventional gilts a real return of 6.6%; treasury bills a 4.7% real return; and inflation-linked gilts a real return of 3.4%.
 * Reasons cited for lagging performance include unexpected rising real interest rates over the 1981 to 2000 period, as well as the fact that high tax rate taxpayers find the bonds' low interest coupons and no capital gains tax attractive, thereby the bonds may be priced to reflect tax factors.

Correlations
The following table provides a matrix of UK asset class correlation over the 1955-1999 period. Nominal return correlations are shown in boldface; real return correlations are shown in standard text. It is important to realize that correlations are not stable, but shift over market environments. Since 1930, the correlation between UK stocks and UK bonds has ranged from .74 (December 1939) to -.31 (May 2007).

Indexing vs. active management in the UK
A white paper published by Vanguard UK Institutional examines the performance of UK active funds against portfolio benchmarks over the fifteen year period, 1996-2011. The paper explores relative performance; the relationship between volatility and return; persistence of active fund performance; market cycle performance; expenses; and performance distribution. A summary of the paper's findings:


 * Returns vs. benchmarks

The paper finds that the median active fund has consistently underperformed benchmarks (as defined in each fund's prospectus), especially when survivorship bias, which has been substantial over the period, is taken into account. The following table provides data for five, ten, and fifteen years. The first percentage shows percentage of active funds under performing benchmark performance (taking account of survivorship bias). The percentage figure underneath shows the annualized excess return of the median active fund (negative returns indicate under performance vs. the benchmark; positive returns indicate out performance vs. the benchmark).

In each of the equity and fixed income categories included in the table above, (for the period 1 Jan 2002 to 31 Dec 2011) the index benchmarks produced higher returns than the median UK active fund with less, or equal volatility.
 * Volatility vs. return

While active strategies in some asset classes showed periods of relative out performance over some bull or bear market cycles, there is no systematic tendency for them to do better at particular stages of the cycle.
 * Market cycle performance

Dividing active fund performance in five year performance quintiles (over the 2001-2006 period), and measuring returns over the subsequent five years, shows that persistence of active fund performance is little different than random. While 16% of top quintile funds (selected in 2006) maintained first quintile performance five years later, the investor faced a 60% chance of falling into the bottom 40% percentile performance or having the fund disappear. Of the 364 funds available to the investor in 2006, only 12 funds (15.6%) maintained top quintile performance over both five year performance periods.
 * Persistence of active fund performance

In general, UK index trackers have lower total expense ratios (TER) than active funds.
 * Expenses


 * Performance distribution

While some UK active funds out perform bench mark UK index funds, the range of dispersion in index fund performance (5 year returns ending 31 Dec 2011) is much lower than the range of active fund performance. Over this period, more active funds under perform the benchmark tracker funds than outperform them, and the range of under performance is generally larger than the range of out performance.