Archive

Posts Tagged ‘outlook’

2022 Year-End Thoughts About 2023

December 22, 2022 2 comments

Use: This article may only be reposted in its unedited entirety (including all links), including the title and author with linkbacks to the original. If you wish to repost in serial form, please contact me via the form at https://enduringinvestments.com to discuss.

When I was a Street strategist, and/or producing ‘sales and trading commentary’ as a trader, it was de rigueur to produce an annual outlook piece. Naturally, everyone does one of those; consequently, I stopped doing them. It seems to me like it would get lost in the shuffle (this is one of the reasons that Enduring’s “Quarterly Inflation Outlook,” which we distribute to customers and is also available by subscription here, is produced on the ‘refunding schedule’ of February, May, August, and November rather than at quarter-end). Having said that – it does seem that, given what inflation has done recently, there are more people asking for my outlook.

I do have to raise one point of order before I begin. As regular readers of this column know, in my writing, I generally try to propose the ‘right questions,’ and I don’t claim to have all the right answers. An outlook piece is often interpreted as being the analyst’s best guess at the answers. While it is that, for me the answers I suggest here are likely to be less valuable to the reader (I do not recommend that you blindly place trades based on my outlook for where markets will go!) than the thought process that is going into them. You may and probably will disagree with some of my answers. But hopefully, you’ll be able to identify where in my reasoning you have specific disagreements, which will either enhance your own view or cause you to thoughtfully reconsider it. That’s the whole point, and I don’t care at all if you disagree! That’s what makes markets.

Moreover…even if my guesses end up being “wrong,” or “right,” based on the actual outcomes in the future, that doesn’t mean they were wrong or right in terms of being a good approach/positioning. Investing is not really all about making the “right” bet in terms of whether you can call the next card off the deck, but about making the “right” bet with respect to the odds offered by the game, and betting the right amount given the odds and the edge. On this topic, I recommend “Thinking in Bets” by Annie Duke as excellent reading.

So, here goes.

MACROECONOMICS

For most of this year, I have been saying that we would get a recession by early 2023. In 2022Q1 and Q2, US GDP contracted. This produced the predictable shrill announcements of recession, coupled this year with sadly simple-minded declarations that the Biden Administration had “changed the definition of recession” by saying we weren’t in one. One television commentator I saw strongly profess the view that the two-quarters-of-negative-growth-is-a-recession definition is “in every economic textbook.” Having read my fair share of economic textbooks and having taught or tutored from a few, I can assure you that is not the case.

I was, and remain, sympathetic to the incoming fire that the Biden Administration took then, because they were basically right: whether we chose to call it a ‘recession’ or not, there was scant sign of any economic distress. Employment (which lags, of course) remained strong, corporate earnings were solid, confidence was reasonably high except for inflation, and citizens still had a substantial cash hoard left over from the COVID stimmy checks. However, while the critics were wrong on the timing they weren’t wrong about the eventuality of a recession. As I also said a bunch of times, there has never been a period where energy prices rose as rapidly as they did between early 2021 and mid-2022, combined with interest rates increasing as rapidly as they did thanks to Federal Reserve policy, that did not end in recession. But it takes Wile E. Coyote some time to figure out that there is nothing under his feet, before he falls, and recessions work similarly. We will have a recession in 2023.

We are already seeing the early signs of this recession. One indicator I like to look at is the Truck Tonnage index, which falls significantly in every recession (see chart, source Bloomberg). The last two months have seen a decline in this seasonally-adjusted index. It is early yet – we saw a similar-sized decline in 2016, for example, so there are false signals for small changes – but the fact that this decline happened heading into the Christmas season gives it more significance.

That’s the goods side. The services side shows up more in the labor market, which lags behind the overall cycle. Yet there too we have started to see some hints of weakness. Jobless claims are well off the post-COVID lows, although they are still roughly “normal” for the tight pre-COVID labor market. And the labor market is really hard to read right now, given the continuing crosswinds from the COVID-period volatility and the fact that so many services jobs now are at least partly virtual. Upward wage pressure is continuing, partly because virtual workers are less productive (shocker reveal there), so this recession in my view will probably not feel as bad as the last couple of recessions (GFC, Covid) have felt. However, we will have a recession in 2023.

The bad news, though is that a recession does not imply that inflation, ex-energy, will decline. Look at this chart, which captures the last three recessions. The post-GFC recession was the worst in 100 years, and while core inflation slowed that was almost entirely a function of the housing market collapse and not the general level of activity. The COVID recession was worse than that, and core inflation accelerated. And the post-tech-bubble recession wasn’t a slouch either; core inflation accelerated throughout 2001 until it started to decline, but only got down to 1.1%, in late 2003.

This chart shows y/y changes, but helpfully shows core-ex-shelter (Enduring Investments calculations). There isn’t a lot to see here in terms of the effect of these three huge recessions.

Lest you think I am just cherry-picking the 2000-2022 period, here is core CPI and GDP normalized as of December 1979. Again, you can see in the GDP line the recessions of the early 1980s, of the early 1990s, and that post-tech-bubble recession. I can’t see those, in the CPI line.[1]

And hey, as long as we are doing this…how about the 1970s malaise when the multiple recessions and flat growth led to … well, not disinflation.

I think the evidence is very clear: forecasters who are relying on the “recession” forecast (which I share) to make a “hard disinflation” forecast are simply ignoring the data. Those two concepts, outside of energy, are not related historically.

That being said, I expect core inflation and median inflation to decelerate in 2023. I just don’t think they will decelerate nearly as much as Wall Street economists think. Shelter inflation is already well above my model, and I expect will come back towards it, but my model otherwise doesn’t see a lot of downward pressure on rents yet. The strong dollar, and some healing of supply chains, will help core goods – but core goods inflation will remain positive next year and probably for a long time, thanks to secular deglobalization, instead of being in persistent slow deflation. And core services ex-rents will decelerate, but mainly because of the technical adjustment in health insurance. Until wages start to ebb, it’s hard to see a crash in core services ex-rents inflation. So that brings me to this forecast for core CPI:

Current2023 Fcast
Core Goods3.7%2.3%
Rent of Shelter7.2%4.8%
Core Services less ROS6.3%5.1%
Core CPI6.0%4.2%

Most of the Street is in the mid-2s for core inflation; the Conference Board forecast for Core PCE recently was raised to 2.8% which would put core CPI at 3% or 3.1%. They’re getting there, but frankly it’s hard to see how you can get to those levels. In my view, most of the risks to my forecast are to the upside.

MONETARY POLICY

An important disclosure should be made here: in 2022, I was utterly wrong about the path the Fed would take. Almost as wrong as it is possible to be. Ergo, take everything I say hereafter in this section with a grain of salt.

Coming into 2022, I thought the Fed would follow the same script they had used for more than a quarter-century with respect to tightening policy: slow, late, tentative, and quickly reversed. Although inflation was already plainly not transitory, I know that the Fed’s models assume a strong homeostasis especially with inflation, to the extent that the persistent part of inflation is essentially (albeit with a lot more math) modeled as a very slow moving average and overall inflation is assumed to pull back to that level. When the Fed talks about the “underlying inflation trend,” that is in simple terms what they are saying. But if you believe that, then there’s very little reason to pursue something similar to a Taylor Rule where policy is driven by simple deviations of growth and inflation from the target levels.

So, when the Fed started to move I expected them to tighten a few times and then to stop and ultimately reverse when financial markets started doing ugly illiquid things. One thing I didn’t anticipate: the markets never really did ugly illiquid things. Investors welcomed the tighter policy, and ran ahead of the Fed to give them room. Especially considering that, at the end of 2021, I think most sophisticated investors viewed the Fed as incompetent (at best) or counterproductive (at worse), the markets gave the Committee an amazing amount of latitude. The Fed, to its credit, saw the gap in the defense and sprinted through it. I did not see that coming.

After nearly 500bps of rate hikes, and a small decline in the Fed’s balance sheet, money supply growth has come to a screeching halt. That’s largely spurious, I think, since money supply growth is a function of bank lending and banks are neither capital-constrained nor reserve-constrained at the moment, and longer-term interest rates have risen but not very much (except in the mortgage market). I suspect that most of the decrease in loan demand that is evidently happening is not in response to the increase in short-term rates but rather to the increase in mortgage rates almost entirely. If that’s the case, then it’s a one-time effect on M2 growth: mortgage origination can only go to zero once. The chart below shows the connection between M2 growth (in blue) and the MBA Purchase index (black). The correlation is not as incredible as it looks, because one is a rate of change that is off-center by 6 months (it’s y/y) and one is a level of activity, but if I expressed both in rate of change you would still say they look suspiciously similar.

If I am right about that point, then the money supply will shortly resume its growth as the overall volume of lending continues to grow without the negative offset of declining mortgage origination. With money velocity on the upswing now, this will support the level of inflation at a previously-uncomfortable level. So what will the Fed do?

Importantly, the Fed won’t really know that inflation isn’t dropping straight to 2% until after the midpoint of the year. But they’ll make the decision to pause rate hikes sooner than that. I think a 5% Fed funds rate is a reasonable target given their assumptions, a key one of which is that if “underlying inflation” is really 2%-3% then a 5% nominal rate will be plenty restrictive.  

What is really amazing to me – which the ‘me’ of 2021 would never have anticipated – is that Fed watchers and market participants are starting to talk as if they believe the Fed might overdo the tightening, raising rates higher than needed to restrain the economy and inflation (yes, I know I said that a recession doesn’t cause lower inflation but it’s an article of faith at the Fed so we need to pretend as if we believe it). It’s incredible, when you think about it: the Fed hasn’t come close to ‘overdoing it’ in a tightening cycle in decades, if by ‘overdoing it’ we mean that they caused a deflationary crash. The Fed has caused plenty of recessions, but core inflation hasn’t been negative since the Great Depression. And we’re worried about them overdoing it?

Naturally, if you don’t think that raising rates causes inflation to come down then any rate hikes at all…actually, any active monetary policy at all…is too much. But in any event, it’s striking to me that the Fed has somehow restored some credibility as a hawkish central bank. Not that credibility per se matters, since expectations don’t cause inflation. But I digress. It’s still pretty amazing.

When Powell was first named Chairman, I was hopeful that a non-economist could help break the Fed out of its scholarly stupor. As time went on I lost that hope, as Powell trotted out various vacuous terms like “transitory” and leaned on discredited models (nevertheless still in vogue at the Fed) such as those which utilize the ‘anchored expectations’ hypothesis. But I have to say, my opinion of him has risen along with the Fed funds rate.

In my view, the biggest Fed error of the last forty years was Greenspan’s move to make the Fed transparent, which caused the pressures on the Fed to be entirely one-way. The second-biggest Fed error follows from that, and that is the tendency to move rates further and further away from neutral, holding rates at such a level by maintaining vastly higher levels of liquidity than were needed to run the banking system. The consequence of this has been a series of bubbles and asset markets at levels where the prospect of future real returns was abysmal. Plus, it led to the heyday of hedge funds where cheap money levered small returns into big returns.

The Powell Fed, for all of its flaws and awful forecasting, has succeeded in getting the yield curve to the vicinity of long-term fair value, which I define as sovereign real rates near the long-term growth rate of the economy (2.00-2.25% in the US – see chart below, source Enduring Investments before 1997 and Bloomberg after 1997). With a Fed inflation target at 2.25% or so in CPI terms, this means long-term nominal interest rates should be in the vicinity of 4%-4.5% over the long term in the context of a responsible central bank. We’re not there, but we’re getting close.

All of which means that I think the FOMC is just about done with hiking rates for this cycle. I believe they will get to 5%, pause, and stay paused for a long time. I do not expect them to lower interest rates, even if there is a recession, unless markets or banks start to have difficulties or Unemployment gets above 6%. That might happen in late 2023, but even if it does I think the Fed will be much more measured about cutting rates than they have in previous cycles. Credit to Powell for the change in attitude.

Those pieces, the Macro and the MonPol, along with my assessment of relative valuations, inform everything else.

RATES, BREAKEVENS, AND CURVES

The long, long, long downtrend in interest rates is decisively finished. As noted above, when inflation is under control and in the vicinity of the Fed’s 2% target, long-term interest rates should be in the vicinity of 4-4.5%. Over the last century, when rates have been away from the 3-5% range it has generally been either because inflation was unstuck on the high side (1970s, 1980s) or unstuck on the low side (1920s, 1930s, 2010s) (see chart, source Federal Reserve and Bloomberg). The long-term downtrend can be thought of as going from unstuck-high inflation, to normal, and overshooting to the downside in the last decade. But we have now definitively ended that low-rates period.

At a current level of roughly 3.5% nominal, 1.4% real, interest rates are ‘too low’ again, but this is normal for an economy headed into recession. Ordinarily, this configuration of events – a Fed nearing the end of a tightening cycle, a recession looming, and interest rates that have risen 320bps over two years – would make me bullish on bonds. And I do think that the first part of 2023 may see a decent rally as the Fed finishes their business and the stickiness of inflation is not yet apparent, but the recession is. Seasonally, you’d really prefer to be long the bond market/out of equities in the last quarter of the year and out of the bond market/long equities in the first quarter of the year, but I think the seasonal pattern will be reversed this year. So we will come in all happy as bond investors, and get unhappy later in the year.

The reason I think the first quarter of the year will be pretty decent for bonds is because of the timing of the recession and of the end of the Fed tightening cycle. But why the selloff as the year progresses? Well, investors will start to see that inflation is not falling as fast as they had expected, the Fed is showing no signs of easing…and the Federal deficit is blowing up.

In FY 2022, the US government had a $1.38 trillion deficit,[2] in an expansion during peacetime. But there are some inexorable effects pushing that higher next year. For example, interest on the debt: higher interest rates will affect only the part of the public debt that has rolled over, but that is an awful lot of it.

In December 2021, the rolling-12-month interest expense on US Debt Outstanding (see chart, source Bloomberg) was $584bln.[3] As of November 2022, the rolling-12-month expense was $766bln. It will be up another $100bln, at least, in 2023. Social Security benefits paid this year were roughly $1.2 trillion, and benefit payments are due to increase 8.7% next year – so, even neglecting the fact that there will be more recipients next year, Social Security should also be $100bln further in the red. That’s $200bln, on top of the approximately $1.4trillion deficit, and I haven’t even considered Medicare, the decline in tax receipts that will occur thanks to a decline in asset markets this year, or the decline in taxes on earned income when the economy enters a recession. A $2 trillion, peacetime deficit is easily in reach and will be much more if it’s a bad recession. The last time we had that big a deficit, the Fed happened to also be buying a couple trillion dollars’ worth of Treasuries. This time, though, the Fed is shrinking its balance sheet.

It is fairly easy to imagine that longer interest rates will have to rise some, in order to roll the maturing debt. As I said, higher interest rates don’t really bother me because I don’t run a highly-levered hedge fund. (But if the rise in rates were to get sloppy or rates were to rise enough to threaten a spiral in the deficit, then I can imagine the Fed stepping in to reverse its balance sheet reduction and being under even more pressure to guide rates lower. However, it’s not my base case.)

Also, as the year goes along the stickiness of inflation will become more apparent and investors will rightly start to put that assumption back into their required return for nominal bonds. One of the really crazy things that happened in 2022 was that inflation compensation in nominal bonds (aka ‘breakevens,’ the mathematical difference between yields on nominal bonds and yields on inflation-linked bonds that pay inflation on top) declined even as the overall level of inflation continued to climb. At the time of this writing, Median CPI has not yet even decisively peaked, although I think it will. But with Median CPI at 6.98%, it’s incredible that the market is demanding only 2.28% annual compensation for inflation over the next decade (see chart, source Bloomberg). That basically says investors are comfortable earning an increment that underpays them for inflation in the near term, and in the long term will only compensate them for what the Fed says they are trying to pin inflation at.

That’s not as easy a trade as it was when 10-year breakevens were at 0.94% in March 2020, but it still seems to me that most of the risk over that decade would be for inflation to miss too high, rather than too low. I understand that the FOMC wants inflation down around 2%. And as for me, I want a Maserati. Neither one of us is likely to get what he wants, just because we want it.

As the first quarter of the year passes and long-term interest rates decline, the curve may invert further from its current level. But I don’t think it can invert that much, which limits the value to being long, say, 10-year notes from this level. Given the current level of inversion, it is fairly easy to construct steepener trades that throw off positive carry. For that matter, a leveraged investor who is financing at 4.5% and earning 3.75% is more likely to want to go the other way! I think it’s going to be difficult to get a good bull market rally going in bonds, and if I was a leveraged hedge fund investor I’d be playing from the short side/steepener side even in the first quarter of the year (albeit cautiously). The chart below (source: Bloomberg) shows 2s/10s monthly going back to 1980. The only time the curve was more inverted was in the early 1980s, a couple of years after Volcker’s Saturday Night Special and with the hiking campaign solidly underway as it is now. I’m expecting 2s/10s to go positive in 2023, although the best shot at something like +50bps would come if the Fed actually did ease. Ergo, a steepening trade is also nice because it works in my favor more if I’m wrong about the Fed staying on hold for a while after they finish hiking to 5%.

Put those together and I see Fed funds at 5%, 2yr Treasuries at 4.25%, and 10s at 4.5%.

We obviously look deeper than that, though, on this channel. We can separate nominal yields into real yields (represented by TIPS) and inflation compensation (breakevens, or inflation swaps). Here are what the curves look like today (source: Enduring Investments).

From here, it looks fairly obvious that a good deal of the steepening should come from longer-term real rates rising. The 2y TIPS bond is at roughly 2%, so 2s-10s in reals is about the same as it is in nominals. The inflation curve is ridiculously flat. I do think that the inflation curve is more likely to shift higher in parallel than to steepen; a steepening inflation curve would imply accelerating inflation going forward and I don’t think investors really believe we’ll get acceleration. So I think that the movement in the shape of the TIPS curve will be very similar to the movement in the nominal curve, but with the level of the nominal curve being driven by an upward parallel-ish shift in the inflation curve.

2y10y
Current TIPS Yields1.96%1.42%
EOY TIPS Yields1.80%1.85%
Current Breakevens2.30%2.27%
EOY Breakevens2.45%2.65%

VOLATILITY

Generally speaking, a higher-inflation environment is a higher-volatility environment. The chart below (source: Bloomberg) shows core CPI in blue against the ICE BofA MOVE Index of fixed-income option volatility. True to form, the higher-inflation regime has correlated with higher levels of fixed-income volatility.

It isn’t terribly shocking that volatility is higher in bonds than it had been during the years when interest rates were fixed within a stone’s throw of zero. And it shouldn’t be terribly shocking that I expect volatility to stay somewhat higher than the 2017-2019 and 2020-early 2021 levels, even as core inflation recedes somewhat. What may be surprising is the observation that a sizeable gap has opened up in the behavior of fixed-income volatility and equity volatility, as the following chart comparing the VIX (equity vol) and MOVE (fixed-income vol) shows. Note that these are different axes, but you can clearly see the uptrend in the MOVE that has not been replicated by the VIX.

I mentioned earlier how regular and controlled the decline in the stock market has been, and how this has allowed the Fed to push rates further than anyone thought they would, a year ago. There have not been too many periods where option sellers have been punished for being short vol in equities. On the other hand, bond vol has been very different now from what it was a few years ago. In short, there has been a regime change in bond vol, but not in equity vol. At some level, this will continue, but the spread should narrow as the Fed gets to the end of the tightening regime. I think we will end 2023 with the VIX above 22 log vol – where it is today or slightly higher – but with the MOVE around 90 norm vol.

Both of those figures represent more-volatile conditions than we have seen for some years pre-COVID.

EQUITIES

It hurts to say, but equities are still far, far, far overvalued.

For many years, there has been a running tension between people who use the “Fed model” as a way to justify the current level of the stock market and the people who point out that the “Fed model” does not imply that the current level of the market is fair. The “Fed model” essentially says that when interest rates are very low, the present value of future cash flows is higher; ergo, the equilibrium value of the average equity (whose fair value is dependent on the present value of future earnings) and hence the overall stock market is higher, when interest rates are lower. This is analytically true. However, it does not mean that your expectation of future returns, when P/E multiples are at 40 but interest rates are low, should be the same as your expectation when P/E multiples are at 15 but interest rates are high. The level of interest rates explains higher equity prices, but it does not imply that those are now long-term fair value levels.

But this tension was almost always resolved in favor of the people who thought that rock-bottom interest rates meant that stocks should be at sky-high multiples, and value investors were left in the dust for more than a decade.

Unfortunately, this tension is being reduced because interest rates are going higher, and may never go back to those levels again. Consequently, equity price/earnings multiples need to re-rate for the new level of interest rates. The same logic that was used to justify the stock market at a 35 Shiller P/E, reconciles to lower prices now and going forward. The chart below (source: Robert J Shiller, updated with Enduring Investments calculations) shows the Shiller P/E (aka Cyclically-Adjusted P/E Ratio, or CAPE) versus 10-year interest rates in the post-WWII period. There is, ex-Internet bubble, a pretty clear relationship between interest rates and valuations. The red dot is where current multiples and interest rates are.

My forecast of 4.5% 10-year Treasuries implies something like a 23 Shiller P/E, down from 30 now. Without earnings growth, that 23% decline in the multiple implies a 23% decline in the stock market from these levels. I don’t think earnings themselves will increase or decrease very much unless the recession is much worse than I think it’s going to be, but the same lag between wages and product prices that flattered earnings when inflation was heading higher will detract when inflation decelerates. Moreover, if I’m right that Powell is intentionally steering interest rates to a level that is consistent with a long-term equilibrium around 4%-4.5% then this 23% adjustment in prices will not necessarily be followed by another massive bull market the likes of which we became accustomed to during the long bond bull market of the last 40 years. A Shiller P/E in the low-20s is still fairly generous historically but it may be sustainable.

So, my point forecast is for the S&P to get to 3,000 sometime in 2023. I don’t think the current bear market will last the entire year, and in fact I am sure there will be a rollicking rally when it is clear the Fed is done tightening. But sticky inflation will hurt here, too, and after that rollicking rally I think we’ll have another low, and from that low is where a modest bull market will begin.

However, I should also note that 1-year equity vol is around 25%, so my projection is within 1 standard deviation of unchanged!

COMMODITIES

From 1999 through 2008, commodities were in a bull market. After a brutal crash in the Global Financial Crisis, commodity indices had another mini-bull market from 2009-2011 before enduring a 9-year bear market. Since March 2020, the massive increase in the quantity of money has driven down the value of money relative to commodities or, to put it in the normal way, has driven up the price of commodities.

The Bloomberg Commodity Index (spot) rose from 59 in March 2020 to 124 in March 2022, and has come off the boil a bit since then. At the highs, though, the level of the index was only back to the levels of 2014. This is normal with spot commodities, which thanks to improved production and extraction technology over time tend to be perpetually deflating in real terms.[4] The good news is that an investor in commodities does not generally buy spot commodities but rather invests through collateralized futures contracts or invests in an index based on collateralized futures contracts. Over time, the collateral return happens to be a very important source of return (in addition to spot returns, the return from normal backwardation, and the volatility/rebalancing return), and this year there is terrific news in that collateral returns are ~4% higher than they were before the Fed started to hike. This means that, all else equal, commodities index returns should be expected to be 4% better (in nominal terms) this year than over the last couple of years. All else is not equal, but I expect gains in investible commodities indices in 2023.

That’s entirely separate from the question of whether we are in a commodity supercycle, due to chronic underinvestment in exploration and extraction technologies and more difficult geopolitical pressures that increase the costs of mining, growing (e.g. because of fertilizer costs/shortages), and transporting the raw commodities. I think the answer there appears to be ‘yes,’ which means that in general I want to play the commodity market from the long side more than from the short side. Of course there will be brutal moves in both directions, and bears will really want to sell commodities as the recession comes to the fore. But most of that is already in the price, with gasoline at levels much closer to the GFC lows than to anything approximating the highs. The chart below shows retail gasoline prices, adjusted for inflation (using 2012 dollars).

Energy prices of course could fall further, but considering that part of the reason prices have fallen this far is that the Strategic Petroleum Reserve has been flushing oil into the system (and that has ended, in theory) and China’s economy has been sputtering under Zero Covid (which has also ended, in theory), it is hard to think that is the better direction at the moment.

OTHER THINGS

I want to append one very important admonition for investors and investment advisors. I mention this frequently on podcasts, TV and radio appearances, at cocktail parties and to random strangers on mass transit:

The next decade will be very unlike the decades we have just experienced. Not only will inflation and interest rates be higher than we’ve become accustomed to, and markets more volatile, but some important drivers of portfolio construction will shift. The good news is that at least some of those shifts are systematic and predictable. The table below shows how 60/40 returns correlate with inflation, with inflation expectations, and with inflation surprise over two periods. The first period was the 30 years ending in 2004, when inflation averaged 4.89% and was three times as volatile as during the subsequent period. During that period, a 60-40 portfolio was significantly exposed to inflation. The more-recent period, during which inflation was low and stable, produced placid 60/40 returns and correlations with inflation that are mostly spurious because there was more noise than signal. Inflation didn’t move!

The first implication of this is that portfolios which have productively ignored inflation-fighting elements over the last two decades need them now, because the main asset classes used in portfolio construction are terribly inflation-exposed. All portfolios for investors who do not have sufficient ‘natural’ inflation hedges should include such assets as commodities and an allocation to inflation-linked bonds in lieu of some of the nominal bond allocation.

The second implication is related but less conspicuous. The entire correlation matrix is shifting away from what it has been over the last couple of decades, and back to something that incorporates the inflation factor that has been dormant. As the most obvious example, stocks and bonds which have been inversely correlated for a while, due to the fact that they respond differently to economic growth, are becoming correlated again. This is not an aberration but entirely normal for regimes in which inflation is not low and stable. The chart below illustrates this. When 3-year average inflation is above 3% (the red shaded area), then 3-year correlations of stocks and bonds tend to be positive (blue line). When inflation is below that level, correlations tend to be negative.

Negative correlations between stocks and bonds are great because they lower portfolio risk. But in the coming decade, 60/40 won’t be as low risk as it has been. But beyond that, the entire covariance matrix that an advisor relies on to simulate and optimize portfolios needs to be examined. The normal way is to use recent returns (say, the last 10 years) to generate this covariance matrix, which then is used to find the mean/variance-optimized portfolio for a given level of risk. That’s normally okay, but as inflation proves sticky that sort of covariance matrix will be wrong, and wrong in a systematic way. What I am doing for our customers is comparing portfolios optimized with a recent covariance matrix to portfolios optimized using a covariance matrix from the 1980s-1990s. It’s important to be aware of this potential problem in portfolio construction, and to get ahead of it.


Finally, let me take a moment to thank the readers of this blog for their interest in it. I write partly because the discipline of arguing my points out thoroughly makes me (I think) a better trader and investor, but I also garner a lot of value from the information and ideas I receive reciprocally from readers who agree or disagree with what I write. I appreciate this feedback very much, and I thank the readers who take the time to share their opinions with me.

Aside from the personally selfish reason I have for writing, there is also the corporate mission the blog is meant to accomplish, and that is to raise the profile of Enduring Investments and the Inflation Guy franchise with prospective clients, and to encourage them to do business with us. If prospective clients see value in these musings, then I hope they will choose to do business with us. Yes, that’s crassly commercial. But ‘tis the season! And if you read this far in this missive, please consider what that means about the value you’re getting, and how much more value you might get from a deeper relationship with Enduring Investments!

And if not, Merry Christmas anyway! Happy holidays and Happy New Year.    

– Mike ‘The Inflation Guy’ Ashton

DISCLOSURE – My company and/or funds and accounts we manage have positions in inflation-indexed bonds and various commodity and financial futures products and ETFs related to them that are discussed in this column.


[1] It bears noting, though, that until 1982 the shelter component of CPI was tied to mortgage rates and home prices and not rents, so that the early-80s rise in core CPI partly reflected the Volcker rate hikes. Fixing that problem was what released the conspiracy nuts who plague us to this day claiming that the BLS “manipulated” CPI downward.

[2] https://fiscaldata.treasury.gov/americas-finance-guide/national-deficit/

[3] Net interest was about $110bln less, since some of that interest is paid to other parts of the government, for example the Federal Reserve system. For now.

[4] I wrote a nice, short little piece called “Corn Prices – Has the Correction Run its Course?” that is worth reading if you are interested in commodities.

%d bloggers like this: