Contagion Risk in Big Bond Bubble

As we roll into another new year, in the AFR I ask: "Is the fixed-rate government bond market – to be distinguished from floating-rate securities – in the mother-of-all bubbles? Very likely. Has it burst? Maybe, although the evidence is not persuasive. As it normalises, will it cause mayhem in other sectors? Probably." Click on that link to read for free or AFR subs can use the direct link here. Excerpt below (and last week's column, entitled "Global boom just getting started", is also here):

"Let's start with a primer. Markets are very focused on fixed-rate government bond yields because these are the "risk-free" interest rates that are used in asset-pricing models – for residential and commercial property, private equity and listed shares – to estimate the present value of the future cash flows produced by the investment in question (eg, a building or company). Recall we need to discount future cash flows because a dollar today is worth more than a dollar in the future given our ability to earn a risk-free return on the dollar by putting it into a government-guaranteed bank account. If you hold your property's rents or projected corporate earnings constant over the next 10 years, and simply increase (decrease) the risk-free interest rate you are using to discount those cash flows back to the current day, the value of your investment will fall (rise). The risk-free rate you select should be maturity-matched to your cash flows, and fixed rather than floating. A floating rate prices off the overnight cash rate and has no long-term interest-rate expectations embedded in it. That is why analysts typically use a fixed, 10-year government bond yield to approximate the market's view of the risk-free rate over the decade in which their cash flows are being crystallised. The average 10-year fixed discount rate in Australia since 1993 – when the central bank started formally targeting inflation – has been 5.5 per cent. The current yield on a fixed, 10-year AAA rated government bond (which is the most common proxy for the long-term risk-free rate in this country) is just 2.75 per cent, or exactly half its historical average. This dramatic lowering of the yield curve (notwithstanding trend economic growth and a skinny unemployment rate) is the key reason why the prices of shares, commercial property, residential property and private equity have all soared since the correction in 2008 and 2009. As governments forced discount rates down by buying their own bonds, the present value of everything automatically inflated. The shortest end of the risk-free yield curve is represented by the Reserve Bank of Australia's overnight cash rate, which has never been lower than its current 1.5 per cent. Since 1993 (2000) the average cash rate has been 4.4 per cent (4.2 per cent). Between the overnight RBA cash rate and the 10-year government bond yield are the market's guesses of the average cash rate over time. It's actually more complex than this. As we move forward along the yield curve, the interest rates we observe reflect both the estimated average cash rate plus a "term premium" that is compensation for the fact that at any particular point you would be committing your money to a fixed-rate bond for that period of time. A rational investor should demand compensation for interest-rate uncertainty over this horizon, especially considering the tremendous volatility in long-term, risk-free rates. One might lock money into, say, a five-year government bond paying a fixed 2.4 per cent interest rate annually only to discover that cash rates rise over this period to, say, 4 per cent. Let's return to the question of the fixed-rate bond bubble. As I noted earlier, the current 10-year government bond yield is 2.75 per cent. That's about 0.40 percentage points higher than its trough in June last year of around 2.35 per cent. So long-term yields have indeed climbed, reducing the value of the associated fixed-rate bonds. But let's put this in perspective. In March 2017, the 10-year government bond yield was 3 per cent, and in June 2013 it touched 4.4 per cent. So even in the recent past, long-term, risk-free rates have been much higher than today's levels. This begs the question: how much higher could they realistically go? The RBA has made it clear that it thinks its normal overnight cash rate should be around 3.5 per cent, or 2 percentage points above the prevailing rate. Remember we also need to add in a term premium for interest-rate risk on future yields. So if we assume that five-year (10-year) risk-free rates ascend from 2.40 per cent (2.75 per cent) today to, say, 4 per cent (4.5 per cent), the value of a five-year (10-year) government bond will shrink by about 8 per cent (17 per cent). That's telling us that a supposedly risk-free, AAA rated Aussie government bond will lose 17 per cent of its value if the 10-year yield simply returns to around its mid-2013 level. That loss has nothing to do with credit risk per se, and wholly reflects the fact that by fixing your interest for 10 years you are making a massive bet on the direction of interest rates over the next decade. Obviously if you don't care about mark-to-market changes in the value of the bond and are happy to hold it for a decade (and never sell it for liquidity), then these movements are irrelevant. Now extrapolate that mark-to-market impact on the risk-free benchmark to risky assets like property and shares. One can imagine potentially larger valuation changes, which could be quite shocking for folks who regard these investments as a one-way bet. The good news is that we are not going to see a big jump in these long-term risk-free interest rates unless we experience much higher consumer price inflation, which would warrant lifting cash rates. Most investors are completely ignoring this possibility. And right now inflation is, in fact, benign. But my base case is that we do get above-target core inflation in the next two to three years. Whether this translates into a sudden or gradual shift in risk-free rates – and the related valuations of all other asset classes – nobody really knows." Click on this link to read for free.