Caveat emptor: triggers linked to inflation rates

Tweet
Before over-engineering the number you wish to impose for locking into your inflation-only hedge you may wish to consider the shortcomings of the measure you are using to compare against the inflation level you have set yourself for hedging. There are two commonly used measures: cash breakevens and inflation swaps. We consider each in turn.



A) Trigger is being compared to cash breakevens calculcated using the Fisher Equation

Devised by Irving Fisher, the Fisher equation links nominal yields and real yields and can be applied to the bond market.

Market shorthand is

n = r + p + k

n = yield on a nominal bond
r = real yield on an index-linked bond
k = inflation risk premium

This  is further truncated to

n = r + bei (breakeven inflation)

This a crude estimate of the breakeven inflation rate for the following reasons:

1. This approximation is better for lower values of n and r so a tick given our current low values of n and r.

2. It is better when the bond market convention is semi-annual rather than annual (the UK is semi-annual so, this works for the accuracy of the approximation).

3. There is usually a term mismatch between the comparator bonds.

4. The bonds are coupon paying bonds rather than zero-coupon

5. The indexation-lag means you aren't really locking into the real yield even if you were to buy the nominal comparator.

6. As time passes since the execution of your inflation hedge (eg buying an index-linked bond and paying away the fixed element), it is debatable as to what real yield you are locking in, by the execution, at a later date, of a matched maturity nominal bond.


B) Trigger is being compared to a market-standard (zero coupon) inflation swap rate

1. Inflation swaps have a standard 2-month lag and so the comments above about lagged inflation also apply here.

2. As time passes since the execution of your inflation (swap) hedge, it is debatable as to what real yield you are locking in by the execution of a matched maturity nominal swap.

For example, assume 5-years pass since the execution of a 30-year inflation swap. The rate achieved on the execution of a matched maturity interest rate swap 5-years later cannot simply be compared to the headline rate achieved on the 30-year inflation swap executed 5-years earlier. An allowance needs to be made for the mark-to-market which has accrued on the inflation swap and, most likely, a real rate will then be calculated as that which was achieved for the remaining 25-year term rather than a real rate achieved for the full 30-year term.

3. Seasonality effects, meaning that, for example it is the case that due to the "seasonal" way in which goods/services are priced eg December sales, impact of VAT increases introduced on a particular date then inflation measured over a 12-month period starting in January could be significantly different from inflation measured over a 12 month period starting in, say, March.

By significant, I mean between 5-10bps. To further confuse us, this is somewhat euphemistically referred to by practitioners as
"seasonal effects"
"technical reasons"
" mechanical impact on RPI of inflation seasonality"


The view from 10,000 feet:

1. Triggers are a welcome addition to the LDI toolkit. They introduce discipline into decision-making.

2. When bifurcating rates and inflation hedging, then thinking carefully about your inflation trigger level does not mean over-engineering your actual inflation trigger rate. Rather, it is more important to recognise that there are embedded approximations in the calculations.

Tweet