Archive

Archive for July, 2020

Low Real Yields – You Can’t Avoid Them

July 29, 2020 7 comments

Recently, 10-year real yields went to new all-time lows. Right now, they’re at -0.96%. What that means is that, if you buy TIPS, you’re locking in a loss of about 1% of your purchasing power, per year, over the next decade. If inflation goes up 2%, TIPS will return about 1%. If inflation goes up 8%, TIPS will return 7%. And so on.

With that reality, I’ve recently seen lamentations that TIPS are too expensive – who in the world would buy these real yields?!?

The answer, of course, is everybody. Indeed, if you can figure out a way to buy an asset without locking in the fundamental reality that the real risk-free rate is -1%, please let me know.

Because when you buy a nominal Treasury bond, you are buying them at a nominal interest rate that reflects a -1% real interest rate along with an expectation of a certain level of inflation. The whole point of the Fisher equation is that a nominal yield consists of (a) the real cost of money, and (b) compensation for the expected deterioration in the value of that money over time – expected inflation.[1] So look, if you buy nominal yields, you’re also getting that -1% real yield…it’s just lumped in with something else.

Well golly, then we should go to a corporate bond! Yields there are higher, so that must mean real yields are higher, right? Nope: the corporate yield is the real yield, plus inflation compensation, plus default risk compensation. Your yield is higher because you’re taking more (different) risks, but the underlying compensation you’re receiving for the cost of money is still -1%.

Commodities! Nope. Expected commodity index returns consist of expected collateral return, plus (depending how you count it) spot return and roll return. But that collateral return is just a fixed-income component…see above.

Equities, of course, have better expected returns over time not because they are somehow inherently better, but because buyers of equities earn a premium for taking on the extra risk of common equities – cleverly called the equity risk premium – over a risk-free investment.

In fact, the expected returns for all long positions in investments consist of the same basic things: a real return for the use of your money, and a premium for any risk you are taking over and above a riskless investment (the riskless investment being, we know, an inflation-linked bond and not a nominal bond). This is the whole point of the Capital Asset Pricing Model; this understanding is what gives us the Security Market Line, although it’s usually drawn incorrectly with T-bills as the risk-free asset. Here is the current market line we calculate, using our own models and with just a best-fit line in there showing the relationship between risk and return. Not that long ago, that entire line was shifted higher more or less in parallel as real interest rates were higher along with the expected returns to every asset class:

So why am I mentioning this? Because I have been hearing a lot recently about how people are buying stocks because TINA (There Is No Alternative) when yields are this low. But if the capital asset pricing model means anything, that is poor reasoning: your return to equity investment incorporates the expected real return to a riskless asset. There is an alternative to equities and equity risk; what there’s no alternative to is the level of real rates. The expected real return from here for equities is exceptionally poor – but, to be fair, so are the expected real returns from all other asset classes, and for some of the same reasons.

This is a consequence, of course, of the massive amount of cash in the system. Naturally, the more cash there is, then the worse the real returns to cash because a borrower doesn’t need to compensate you as much for the use of your money when there’s a near-unlimited amount of money out there. And the worse the real returns to cash, the worse the real returns to everything else.

You can’t avoid it – it’s everywhere. I don’t know if it’s the new normal, but it is the normal for now.


[1] Unhelpfully, the Fisher equation also notes that there is an additional term in the nominal yield, which represents compensation being taken on by the nominal bondholder for bearing the volatility in the real outcome. But it isn’t clear why the lender, and not the borrower, ought to be compensated for that volatility…the borrower of course also faces volatility in real outcomes. In any event, it can’t be independently measured so we usually just lump that in with the premium for expected inflation.

Summary of My Post-CPI Tweets (July 2020)

Below is a summary of my post-CPI tweets. You can (and should!) follow me @inflation_guy. Or, sign up for email updates to my occasional articles here. Investors, issuers and risk managers with interests in this area be sure to stop by Enduring Investments!

  • Welcome to CPI day from the Rocky Mountains! Inflation indicators are getting more and more important these days, and today’s number will be interesting as always.
  • First, let me mention that I will be on @TDANetwork with @OJRenick today at 9:15ET to discuss the CPI and other inflation items. Tune in!
  • In this month’s CPI, we will probably continue to see some of the data collection issues from the last few months, but fewer. As the US has opened back up, the readings should start being cleaner. So the volatility we see will start to represent actual volatility in pricing.
  • I think we should continue to be cognizant of the underlying context, and that is that M2 growth in the US is at all-time records. Can we have 23% money growth and no inflation?

  • Keynesians would say yes – in fact, with the massive output gap, we should be in outright deflation. There may be some nuance (inflation anchoring keeps it from happening immediately), but the sign is clear.
  • Monetarists would say that after velocity corrects the unusual, huge precautionary demand for cash balances (lowering velocity), we should see problematic inflation. How much? Hard to say, but I’ll say I’m not aware of a society that has had 20% m2 growth and NOT had inflation.
  • So far, the monetarists are winning. While core inflation has dipped the last few months, it’s all in the ‘left tail’ of the distribution: used cars, airfares, lodging away from home, and apparel account for it all. Median inflation has barely budged.
  • So the story going forward is that we are going to eventually see a bounce in those tail items, so if we’re going to get deflation we need to see broader deceleration in prices.
  • The consensus today is for 0.1 m/m on the core CPI. That seems low to me. Unless something weird happens with primary rents and OER and they start to slow markedly, any rebound in those tail items is going to take inflation up. But we will see, in just a few minutes. Good luck!
  • Higher than expected core: +0.19% m/m, +1.19% y/y. As I noted up top, that’s not super surprising actually.

  • Sorry – erratum: seasonally adjusted core was +0.235, not 0.19%. So almost gave us a double-tick surprise.
    • In some of the unusuals from past months, yes we’re seeing corrections: Airfares +2.60% m/m (last month -4.88%); Lodging Away from Home +1.21% (last month -1.53%). Used Cars still soft, -1.19%. That’s actually surprising.
  • I suppose it is just in the lag, but used car prices rebounded in a major way in the private surveys (see chart). the Black Book index is higher than before Covid.

  • Interestingly, the upside core surprise happened DESPITE softness in primary rents (+0.12%, 3.22% y/y vs 3.48%) and OER (+0.09%, 2.84% y/y vs 3.06%)
  • Seeing strength in medical care, not surprisingly. That’s been hard to survey over the last couple of months. Doctors +0.45% m/m, 2.10% y/y (was 1.80); Hospitals +0.37% m/m, 5.29% y/y (was 4.86%) Pharma still soft, flat this mo but y/y up to 1.38% vs 0.97%
  • Forgot to mention Apparel, that other one-off. +1.66% this month, though still down hard y/y.
  • Hospital Services y/y

  • Core ex-housing rose slightly, to +0.35% y/y from +0.29% y/y. Still very low, but I suspect the lows are in.
  • Although this is an upside surprise, core still fell vs last month on a y/y basis because we rolled off a +0.28% from last June. Core goods were -1.1% vs -1.0% last month, core services +1.9% vs +2.0% last month (y/y). So there’s still some downward pressure. But it’s turning.
  • I’m afraid this is going to be abbreviated this month as I need to go get ready for @TDANetwork. I’ll be on at about 9:15 ET. Later I’ll update the four-pieces charts.
  • Bottom-lining it: the monetarists win another round as the broad deflation Keynesians would predict is nowhere in evidence. Some of the one-offs are correcting. We do need to keep an eye on the softening in rents.
  • NONE of this will bother the Fed yet…I doubt any of them got out of bed for this number. Thanks for tuning in!

Today’s CPI examination is somewhat abbreviated as I am on ‘vacation’ and my inspection of the report was interrupted by my TDA Network appearance. We’ll be back to normal-length next month. The basic story for the June CPI core inflation figure is a bit of a return to the mean, with most of the left-tail items (with the exception, only temporary, of used cars) rising and retracing some of the dramatic declines we saw in March, April, and May and at the same time a bit of softening of rents. I think the rental softness is also temporary, and tied to a lack of traffic over the last few months – but that’s the only potential fly in the ointment.

The takeaway for this month is really provided by the chart of the last 12 months’ core CPI print, the second one above. Heading into COVID, core CPI was printing around 0.2%-0.25% fairly consistently. March, April, and May were deep exceptions, but June is back to the former trend. To be fair, median CPI this month was very low, at +0.12% m/m and “only” 2.60% y/y. That deceleration was tied to softness in South and West regional Owners’ Equivalent Rent indices, which as I said would surprise me if it was repeated. Rents had recently retreated to our model, and have now slipped below (see chart below).


So, is a dramatic acceleration in inflation evident in the data? Not yet, although the acceleration in M2, coupled with my conviction that the precautionary demand for cash balances will eventually relax, makes me confident that we will see that in the quarters to come. But if we can’t say that breakout inflation is here, neither can we say that there is the least sign of deflation in these figures. Again, if the Keynesians are right, then the huge output gap means that aggregate demand will be insufficient to keep prices up, and deflation will ensue. In the Keynesian world, the only reason this hasn’t happened yet is that inflation expectations are well-anchored or, in other words, merchants haven’t caught on yet that they can’t charge what they were previously charging. In the monetarist mindset, the fact that there is far more money in bank accounts than there was just a few months ago means that eventually the nominal price of goods will be bid up since the price is after all just an exchange rate (of dollars for stuff), and when there are many more dollars then the price of those dollars declines (the price of stuff rises). The reason inflation hasn’t happened yet, according to this view, is that people are clutching nervously onto those cash balances, rather than spending it.

But there are signs that nervousness is ebbing – such as in the price of automobiles and many online goods. If the country lurches back into lockdowns, and the recovery turns back into a deeper recession, then that nervousness may continue for longer. American consumers, though, suck at saving and big cash balances and high savings rates are not usually persistent.

Categories: CPI, Tweet Summary
%d bloggers like this: