Wednesday 24 March 2021 by Jonathan Sheridan long-dated-main Trade opportunities

Long bonds and rising yields

I have had more than a few conversations in the last week or so with clients with the general theme of being nervous about buying longer dated bonds given the move in government bond and swap rates since late last year.

There are 3 points I would make about the move:

1. Rates are still at historically low levels

Look at the longer term chart below and the trendlines – we haven’t broken out of the long term downtrend which in my view is the overarching issue – we have structural changes implemented in the last 20 years or so which are driving yields lower: debt, demographics and technology. These 3 megatrends are also likely to drive deflation over time and do not reverse easily, even in the face of monetary and fiscal stimulus of a size not seen since WW2.

long-dated-1

Now I don’t have a crystal ball, and so I can’t say for sure that the current reflation narrative will end up in real inflation or if it will peter out as the stimulus recedes. I do know for sure that we (the developed economies of the world) cannot afford to keep this level of stimulus in place for an extended period, and that people have quickly forgotten the shock to global growth that COVID was.

We won’t be back to trend growth (which has been slowing for decades) for at least another year and that is with the support of this huge stimulus.

WEB-banner-webinars-Generic

Look at the even longer trends:

long-dated-2

Do I even care?

A little flippant so please don’t take it to heart, but the reality is most portfolios are of short duration due to the way we advise on construction. This includes a (usually) healthy allocation to both floating rate notes and high yield credit, and even the longer bonds are in credit – they aren’t pure interest rate exposure via government bonds. Even if they are then the portfolio weighting is usually small – below 5% - so a move in price doesn’t affect the portfolio return materially, and in absolute terms might even be only a rounding error.

I had a look through all my client portfolios yesterday and the average duration – the sensitivity to a 1% move in yield – was in the low 2s range. This means for a 1% move in yield the portfolio as a whole will move by ~2%. To compare this to equities, the US VIX at 23 is currently pricing a daily move in the S&P500 on average of approx. 1.5% and the Australian equivalent is at 1.1%.

It took the bond market from mid-October to mid-February to move 1% up in yields which would have moved the average portfolio value 2% - and that was in the quickest steepening of the yield curve in 20 years.

Bond volatility is very low if you are invested in credit, which is what we overwhelmingly are. Credit spreads usually counteract a movement in government rates, resulting in only a small change in the actual price of the bond.

3. Steep yield curves are good for longer dated bonds

This is just basic mathematics, but the concept of “rolling down the curve” is not one usually appreciated outside bond market participants.

Take this simple, hypothetical example.

I buy a 10 year bond at 3% yield and coupon for a price of 100. The 7 year yield for the same issuer is 2.50%.

Therefore in 3 years when that 10 year bond is now a 7 year bond, all other things being equal, the yield on the bond will be 2.50%. For this yield to be achieved, the price of the bond must be 103.5:

Price   103.5
Coupons 1 3
Coupons 3 3
Coupons 4 3
Coupons 5 3
Coupons 6 3
Coupons 7 3
Maturity 100
Total return 17.5
Per annum 2.50%

At a price of 103.5, the return to the investor who bought the 10 year bond at 100 with a 3% coupon will be 4.17%.

For the price to be 100, i.e. no mark to market loss, the yield on the 7 year bond would need to be 3% and so the 10 year yield at the time would need to be 3.5% assuming the shape of the curve remains the same. Both of these would be considered big moves.

Remember as I said above, credit spreads also play into the stability of the bond price. For the price of the bond to fall, the overall yield of the bond has to rise. Credit spreads are the market’s extra yield required to take the credit risk of the issuer compared to the risk-free rate. If growth and inflation are picking up, this is typically a better environment for credit and therefore the credit spreads will typically reduce when base rates rise, offsetting each other.

As there is more risk the longer you go, credit spreads also typically reduce at a faster rate than risk free rates for a reduction in tenor, i.e. going from a 7-10 year maturity.

Conclusion:

I don’t know where 10 year rates are going from here, although the sentiment certainly seems for higher rates. Remember though that most people suffer from recency bias – what has just happened is more likely to be believed than something that happened a longer time ago.

We have not dealt with the debt overhang from the huge amount of indebtedness that existed prior to COVID, and that has just got worse as everyone – corporates and governments – have borrowed hugely with rates low to get the liquidity to survive the COVID recession. Debt is tomorrow’s consumption brought forward to today, so by definition more debt today means lower growth in the future. Read Lacy Hunt’s work for the economics behind the debt multiplier.

I am confident in the ability of credit – even 10 year credit – to roll down the curve and deliver expected, and hopefully higher than expected returns to holders. Remember a 10 year bond investment is exactly that – it runs for 10 years, and that should be the expected holding period for the bond unless a more favourable situation presents itself.

In Australian investment grade credit you have a historical default risk of almost zero, so your only real potential for loss will be an opportunity cost loss, which can be dealt with by having mental fortitude and the ability to look across the whole (multi asset not just bond) portfolio and look at risk adjusted returns. I am also confident in my principles of portfolio construction, which is proved out by the low duration of the vast majority of client portfolios.

If a 10 year bond looks like good relative value the chances are it will return more than the coupon rate or yield at purchase unless we really enter a cyclical shift of inflation which is genuinely growth driven. My personal view is this is unlikely but I could always be wrong, hence I don’t have all my eggs in the one basket.

Hamish Douglass, founder of Magellan, made this comment recently:

“If I have to take a view I think this will be transitory, the fiscal stimulus will pass through the economy, and then we are going to be looking back into a factual situation of lower long-term structural economic growth.,”

I would sit in this camp but am happy that even if rates rise more quickly than expected then portfolios will be well positioned to take advantage as shorter dated bonds mature and can be reinvested at those higher rates.