It is often quoted that the main risk that bond investors face is that of inflation. This is broadly speaking true. A bond investor purchases an asset with known forward stream of cash flows. If the purchasing power of those cash flows is eroded over time, the investor will lose out.
One might argue that this is equally true of equity and property investments, but these types of investments have a degree of natural hedge against inflation typically dividends and rental incomes rise with inflation over time.
From the point of view of pension fund managers and trustees, hedging against the risk of inflation has always been one of the main priorities. Funds of this nature have long-term liabilities, often linked to inflation or earnings and thus have a strong appetite for inflation-linked assets to put on the other side of the balance sheet. Given the demand from the UK’s pension fund industry, the creation of index linked bonds was not surprising. Perhaps what was surprising was who was the first government off the block to issue them none other than our own HM Treasury. 1981 saw the first issue of UK index-linked gilts (commonly known as “linkers”) hit the market. Amazingly, the more sophisticated US Treasury market did not follow suit with its own index-linked securities (Treasury Inflation Protected Securities or TIPS) until 1997.
Let’s take a look at that first index linked security - £1 billion of the Treasury 2% Index-linked 1996 was issued on the 27th March 1981 following the announcement of this new class of security by the then chancellor Geoffrey Howe in the previous budget. Purchase of the instrument was restricted to pension funds and the like, although this restriction was sensibly removed the next year in order to improve the marketability of the security.
So how did this security work? The basic principle was very simple. All future cash flows (both the semi-annual coupon and redemption payments) were to be linked to the UK’s Retail Price Index (RPI). Thus, should the value of the RPI double by the bond’s maturity date, the holder will receive coupons at the rate of twice the initial 2% coupon and a redemption payment* of twice the “par” (100) redemption price.
The technicalities were a little more complex than conventional gilts. It was important to ensure that the next payable coupon was a known quantity in order to facilitate secondary market trading (particularly the calculation of accrued interest). In order ensure this, an eight-month lag was applied to the indexation process. All subsequent payments were then adjusted to by the ratio of the 8-month lagged RPI to the “base” RPI, the latter being determined by the value of the RPI 8 month’s prior to the launch. This system worked well enough, although cynics will point out these bonds offer no protection against inflation for the last eight months of its life.
If we take a look at the Treasury 2% 2006 Index Linked gilts, issued in July 1981, here are the actual payments received by investors holding the bond from issue to maturity, based on a £100 nominal value.
The first point to note is that the first sum received of 92p is less than the 100p one might expect from a semi-annual payment of the 2% coupon. My first assumption was that this was due to a short-lived spell of deflation. However this was not the case. The first coupon was fixed at the time of issue and took account of the fact that the bond was part-paid in three installments (March, August and September), reflecting the delayed investment process.
From then on, the process was straightforward. The “base” RPI was set from data calculated eight months before the issue. In the case of this partly-paid issue the figure eight months prior to July was applied, this being 274.1 on the 1975 rebased data*. The amount payable for the second coupon in July 1982 was then calculated by applying the appropriate RPI data point, in this case the RPI eight months before July 1981 and comparing it to the base. Thus, for £100 nominal of the bond:
RPI (July 1981 - eight months) X semi annual coupon (1%)
Or 306.9 x 1% = £1.11 (coupon payment are rounded down to the nearest penny)
And so on and so forth. Each semi-annual payment is adjusted for the RPI. This was almost always upwards but I note that July 2002 saw a very small downward move. Illustrating the point that indexation on these bonds is not “upward only”.
Finally, the holders received the maturity payment of £278.63, some 14 years after the bond launched and in 2006 the world’s first index-linked bond was retired. How good did this protection against inflation prove to be? Certainly, the holders enjoyed the security of quantifiable hedge against the RPI but it might be worth considering that conventional gilts offered yield of 9-11% in the early 1980’s, a figure that would have considerably outperformed the total return seen on our index linked bond.
Consider also that the so-called inflation link may not have been perfect. If I remember correctly a pint of beer could be had for about 65p in most areas; perhaps 95p in a more fashionable establishment. By 2006 I’d say that £3 was the norm.
Moving back to the world of bonds, numerous “linkers” followed the first issue, with the new asset class finding strong demand from the pension fund industry. Many examples of these “mark I” index linked are still trading in the markets and their relatively simple calculation basis makes them a popular choice. At the time of writing, the following index linked gilts were available for trade on the Bondscape platform. Note the high prices of these instruments, a result of the fairly distant issue date and the subsequent rise in the RPI.
UK Index Linked Gilts
|GBP||12 May 2010||UK Gilt I-L Stk||GB0009063578||2.5||23 Aug 2011||1 yr 3 mths||309.33||0.73|
|GBP||12 May 2010||UK Gilt I-L Stk||GB0009036715||2.5||16 Aug 2013||3 yrs 3 mths||272.255||1.94|
|GBP||12 May 2010||UK Gilt I-L Stk||GB0009075325||2.5||26 Jul 2016||6 yrs 2 mths||304.305||3.15|
|GBP||12 May 2010||UK Gilt I-L Stk||GB0009081828||2.5||16 Apr 2020||9 yrs 11 mths||307.25||3.69|
|GBP||12 May 2010||UK Gilt I-L Stk||GB0008983024||2.5||17 Jul 2024||14 yrs 2 mths||270.21||3.88|
|GBP||12 May 2010||UK Gilt I-L Stk||GB0008932666||4.125||22 Jul 2030||20 yrs 2 mths||257.975||3.85|
|GBP||12 May 2010||UK Gilt I-L Stk||GB0031790826||2.0||26 Jan 2035||24 yrs 8 mths||157.825||3.82|
Since the 1981 launch of the original UK index-linked gilts, various developments occurred in the international markets, notably the emergence of the US Treasury Inflation Protected Securities or “TIPS”. The US had followed a model established by the Canadian government in the structure of its index-linked bonds and this gradually become the norm in the international capital markets and in 2005 the UK government followed suit ( as indeed did many other governments) with the establishment of “Canadian style” index linked gilts.
These new type of linkers were not, as many gilts dealers expected at the time, issued with a side of maple syrup. The “Canadian style” structure referred to a three-month lag structure for the calculation of coupons and redemption amounts. Arguably, this is a fairer method of calculation, given that the shorter lag means a tighter relationship between the index and the security. However, it does make the calculation of the accrued interest from the six-monthly coupons fairly tricky. An estimate has to be made, using the “index ratio” a factor published daily by the DMO.
The next complexity for investors to get their heads round is the pricing convention. “Old-style” linkers trade on an inflation adjusted basis. As we can see from the table above, this means that the cumulative effect of inflation has increased the secondary market price of these bonds to two or even three times their original issue price. However, the “type two” or “Candian-style” linkers trade on what is known as a “real price” basis. This means that the inflation has been stripped out and as a result prices tend to gravitate around par. However, when the bonds are settled, the consideration will reflect the accrued inflation. The “inflation adjusted dirty price” is calculated as follows:
Inflation adjusted dirty price = (Index Ratio x real clean price) + (Index Ratio x real Accrued Interest).
At the time of writing, the table below shows the “type two” linkers trading in the market with the respective clean and dirty prices. Again, thanks to Winterflood Securities for the provision of this data.
|Price without inflation
||Price with inflation
|Yield with inflation assumption at 5.18%
||DMO Real Yield
|Treasury 1.25% 20017||107.25||123.50||0.199||5.438%||0.257%|
|Treasury 1.25% 2027||105.2||120.93||0.199||6.121%||0.920%|
|Treasury 1.25% 2032||108.47||111.44||0.178||6.037%||0.837%|
|Treasury 1.125% 2037||109.44||120.72||0.172||5.935%||0.738%|
|Treasury 0.625% 2040||97.30||100.25||0.289||5.924%||0.727%|
|Treasury 0.625% 2042||98.56||100.25||0.091||5.872%||0.676%|
|Treasury 0.75% 2047%||104.54||103.48||0.112||5.804%||0.609%|
|Treasury 0.5% 2050%||96.72||101.19||0.159||5.793%||0.599%|
|Treasury 1.25% 2055||128.22||148.82||0.201||5.742%||0.548%|
On the subject of “new” linkers vs “old” linkers, there is one observation to be made. It is notable that the fixed element coupons of the more recent issues are considerably lower than the 2.5% seen on the original 1980’s issues. This perhaps reflects the generally lower expectations of investment returns of the modern world
Estimating value in index linked gilts
With conventional bonds, an accurate assessment can be made of the forward yield of the asset. Future cash flows are known and investors can lock in a yield to maturity based on the current market level.
Not so index linked bonds. Both the future coupons and the future redemption price are unknown, so how can value be established? The best one can do is to make an educated guess. An assumption (that most dangerous of concepts!) must be made as to the future rate of inflation. One technique to apply is the calculation of the “money yield” of the bond. To perform this calculation, the “core” or average rate of inflation is applied to the future cash flows of the bond, and this in turn can be discounted back to calculate a conventional YTM. Another popular tool for valuation, or at least comparative judgment is the “breakeven inflation rate”. This data is published in the FT and other financial publications and shows the inflation rate at which index-linked bonds will “break even” with conventional gilts of an equivalent maturity. This is handy tool. Typically a buyer of index linked bonds will have a good idea of what level of inflation he expects or needs to beat and this concept will tell him or her whether he'd be better off with conventional or index linked assets over the given scenario.
Perhaps the next technique for valuation is the “real return at X% inflation”. With this technique a preset assumption for inflation is applied to the bonds, and the results calculated in “real” or inflation-adjusted terms. The table below is provided by market maker Winterflood Securities and shows this type of calculation based on an assumed 3% rate of inflation over the various periods. It is interesting to note that in some cases the real rate of return is negative. Of course, the end result may differ very greatly from this projection. Should inflation come back with a vengeance, the bonds featured will show considerable growth in both coupon and redemption value over the holding period.
At the time of writing, the new Conservative/Liberal coalition government is attempting to reduce the enormous pension fund liabilities of the public sector. One technique to tackle this is to move the indexation benchmark for such pensions from the RPI to the CPI, the latter having been a historical underperformer. What is good for the goose is good for the gander and should the public sector shift to CPI indexation, it is likely that many private sector defined benefit schemes will also move to this slightly less arduous benchmark. This will create a curious situation of fund managers attempting to balance future CPI-linked liabilities with RPI-based assets. No doubt, ingenious investment bankers will come up with some solutions, but from a longer-term point of view, the obvious solution is for the government to commence issuance of CPI-linked gilts.
Index-linked gilts, yes or no? The natural buyer of index linked gilts is a large pension fund, the trustees of whom need to match long-term index-linked liabilities. Such investors are prepared to give up a degree of performance in the security in return for a government-guaranteed hedge against inflation. Thus, all things being equal, one might expect index-linked gilts to underperform other more risk-positive assets over time.
This is a fair assumption, however, with index-linked gilts, as with most other asset classes, every dog will have its day. An investor or trader purchasing inflation-linked securities during a period of low inflation expectations may be pleasantly surprised by the performance of such bonds when those expectations shift. “Linkers” are not one of my favourite types of bonds, but they are undoubtedly worth monitoring as the performance of the UKTI 2.5% 2020 over the 2009-2010 period demonstrates in the chart above.
Stockcube Research, July 2010