How we know QE has created a bond bubble
Inflation expectations derived from the difference between conventional and inflation-linked gilts do not agree with inflation expectations derived from the yields of each. That shouldn't happen in rational markets.
The yield difference is the true 'crowd' forecast: above-target inflation. The forecasts implied by absolute yields independently are of deflation. But there is no other evidence this is what the crowd believes will happen. Absolute yields bear no relation to market expectations. They are a bubble created (unusually) not by markets but by the Government.
ILG yields are priced off nominal bond yields via arbitrage by investors who hold, and can express in their portfolio selection, views about future inflation. Not all investors conduct this arbitrage. Many are forced to hedge liabilities for accounting or regulatory reasons, such as life funds with an annuity book or Defined Benefit pension funds with funding deficits. It makes them largely indifferent to the cost of hedging and, if they have real liabilities and hold inflation views, they cannot express those views: they have to hedge. We're the other way round: we never hold nominal bonds, only ILGs. Our clients have real liabilities, we see no basis for forming inflation views, inflation is an impossible process to model and we can see that taking inflation risk has historically not been rewarded with a risk premium. But there are probably enough investors in the market who are free to arbitrage the terms and so create a rational-expectations link between the two assets. The difference between the two yields for any one duration is effectively the market’s implied inflation forecast for that time period. At 5 years, for instance, it’s currently around 2.7%. That fits quite well with other evidence of what the crowd believes.
If you only had one instrument, a nominal bond, and so no arbitrage with ILGs, you would have to say that nominal yields such as we have today of close to zero, negative after the trailing inflation rate, were effectively a forecast of deflation, because that’s what would have to happen for the yield to be ‘rational’. Theoretically, you cannot have a bubble in nominal bonds if the yield could always be validated by inflation expectations. But we know independently that deflation is not a true crowd view so it cannot validate the level of nominal yields.
If you only had one instrument because we lived in a 'perfect' world free of money illusion and all bonds were inflation-indexed, yields would not be affected by the force or direction of price changes in the economy (as long as the indication was symmetrical, in both directions). Real yields would instead fluctuate around the pure time value of money plus a small premium for government credit risk. Normal yields, whatever was happening to prices, might therefore be about 1%, and certainly always positive. Not -2%, like today's (see chart).
Hence we can be sure that the implied inflation derived from the difference between nominal bonds and ILGs is an anchor (it’s a real crowd forecast) and the levels of nominal and real bond yields are both distortions, showing up as a difference in implied forecasts of inflation. Only QE can explain this difference. It wouldn't arise with market agents, even with the presence of the Government as issuer, and even with many of the agents in the market being effectively forced to hedge.
A bond bubble sits like a single over-arching idea linking a series of current investment topics:
goal-based asset allocation policy
tactical asset allocation
'false' risk categories for products and portfolios (because bonds are used to locate them on a risk spectrum defined by volatility)
DB to DC pension transfer terms (because cash transfer values are priced off gilts)
the economic effects of rising DB pension deficits
whether QE is doing more harm than good.
The only thing a bond bubble tells us nothing about is inflation.