October 14, 2020

Last week, John Mauldin shared David Rosenberg citing results from various polls (Mauldin’s emphasis):

“All these polls say basically the same thing—it will not be ‘business as usual,’ as the bulls will try and convince you, and the best we can hope for is a partial recovery. I mean, at best. What we had on our hands was a vertical down economic decline with job losses an order of magnitude higher than anything we have witnessed since the Great Depression. So, even as the stock market is telling you it has it all figured out, I can assure you that what we face at this very moment is a very uncertain economic future. And unfortunately, most of the longer-term risks are to the downside.

“We are in a depression—not a recession, but a depression. And I think the dynamics of a depression are different than they are in a recession because depressions invoke a secular change in behavior. Classic business cycle recessions are forgotten about within a year after they end—the scars from this one will take years to heal.”

John concluded with Trump-like emphasis:

Know this: That which can’t go on, won’t. We can’t keep piling on debt at this rate forever, and we can’t repay what we have.

I predict an unprecedented crisis that will lead to the biggest wipeout of wealth in history. And most investors are completely unaware of the pressure building right now.

Never mind the polls, employment stats make the case for a potentially durable crisis:

• The U.S. lost 22.2 million jobs in March-April and only half have been recouped by September. The number of working Americans is still down 6.4% YoY, worst than any other time since WWII.

• In September, 56.6% of U.S. citizens were working, the lowest level since 1975.

• 3.8 million jobs have been lost permanently, and counting as second Covid-19 waves are hitting.

Depressing numbers, depression numbers indeed.

Yet, the broad Wilshire 5000 and the large-cap S&P 500 are above their February highs, investors seeing nothing but blue skies ahead.

Here’s a transcript of Professor Jeremy Siegel’s explanation to Barry Ritholtz in mid-May.

(…) The big difference this time is not only has there been a huge increase in the balance sheet again of the Fed but to a much greater extent, this money is going right into checking accounts, right into transactions account, right into payroll accounts, right into the bank account of individuals and businesses in a way that I’ve never seen before (…).

Those accounts in the eight weeks after the virus hit from the middle of March, the next eight weeks, increased by almost 25 percent. I’d never seen that before.

In the entire year that followed the Lehman crisis, the increase in the M1 money supply was 15 to 20 percent and that’s in a year. (…)

I was privileged — my first teaching job after got my PhD at University of Chicago and I was, as we talked about earlier, a colleague of Milton Friedman and I remember him saying to me, he said, excess reserves are good, it’s good stimulus for the economy but if those excess reserves get pushed in either M1 or M2, they’re going to be far more potent, far more potent, and that is exactly what is happening this time that did not happen last time.

And I think that as we get therapeutic vaccines, as our economy opens up, this liquidity that is in this economy, the Fed is not going to get rid of it. I mean, they basically committed to zero rates and if the government is not going to put a tax increase, this absorbs all this.

I think we’re going to have a huge spending boom next year and I think for the first time, and I know this is a sharp minority view here, for the first time in over two decades, we’re going to see inflation.

In other words, the government, borrowing from the obliging Fed’s printers, sent $1200 checks to people, boosting money supply while people were confined in their home, additionally saving from unavailable services. Unable to travel and dine out, people splurged on their well-being at home, generally unworried about their future income. Meanwhile, the Fed printed more money to purchase just about every kind of traded fixed income instruments.

The chart below shows how the quick and violent drop in spending in March-April reversed as soon as broad money supply started to explode, boosting disposable income and spending on durable goods almost in line with the trend in M2.

Notice that Durable Goods Expenditures are right scaled while M2, Disposable Income and Total Expenditures share the same left scale. Actually, consumers spent a lot less than what landed in their bank accounts. Some paid off their debt, some saved the windfall (both resulting in a higher savings rate) and many others played the stock market.

Obviously, government direct stimulus money (more appropriately called rescue money here) was a big factor in boosting goods consumption, retail sales and equity markets. But not total expenditures.

The next chart shows the tight relationship between aggregate weekly payrolls (employment x hours x wages) and total consumption expenditures. Episodes of sudden bursts and drops in disposable income, in 2008, 2012, 2013 and again in 2020 did not translate into similar changes in spending, suggesting that consumers tend to align their spending with labor income and to save most of windfall monies.

The relationship between labor income and total spending is holding during this crisis. Post the initial shock, total spending is closely tracking labor income, not disposable income. So far, most of the rescue money actually went into savings, not in the real economy.

Since retail sales only account for 33% of U.S. GDP and total consumer expenditures are 68% of GDP, the fate of the U.S. economy still rests mainly on future trends in labor income. Aggregate weekly payrolls were down 1.6% YoY in September, in spite of the 6.4% drop in the number of employed people (red bar). The main offset was wages which are growing 4.7% YoY whereas they were rising 3.0% pre-pandemic. The reason is mainly statistical: most of the job losses since February were in lower wage brackets, boosting the recent averages.

Crucially, in my view, is that employment is swooshing, almost flattening, still down 6.4% YoY in total and 6.8% for private employment, right when many states and cities are imposing new restrictions to control an apparent second wave.

                       Number of employees                        YoY Change

  

Nearly seven months after the lockdowns, with thousands of small and mid-size businesses closed or on the brink, the economy is not on safe and solid grounds as we enter the most important spending season of the year.

The hope is that the current high savings will get used to sustain consumption absent more stimulus, lower unemployment and an immediate vaccine or cure.

Personal savings ballooned from $1.2T in Q4’19 to $6.4T in April and dropped to $2.4T in August as rescue money got spent for survival for most and to splurge or speculate for others. At 50% of the average depletion of July (-$164B) and August (-$724B), savings would decline another $0.9T through December, to $1.5T. If American consumers aim to only spend the same nominal amount as in Q4’19, they would need to dip another $221B in their savings in Q4 which would bring the savings rate to about 8.3% from 7.3% in Q4’19. Each additional 1% growth in expenditures would require another 1.0% dip in the savings rate.

In other words, just to grow total nominal expenditures by 1% YoY, the savings rate needs to quickly come back to its pre-pandemic level.

During 7 of the last 8 recessions, consumers actually increased their savings during and after the recessions. Following the Great Financial Crisis of 2008-09, scarring effects made people actually maintain their savings rate at a higher level than during the previous decade, something that could very well happen this time around for obvious reasons.

The consumer sector could thus be in for very difficult times and betting on strong consumer spending in 2021 is like defying the (G)odds.

Something Jeremy Siegel is willing to do. To repeat from above:

And I think that as we get therapeutic vaccines, as our economy opens up, this liquidity that is in this economy, the Fed is not going to get rid of it. I mean, they basically committed to zero rates and if the government is not going to put a tax increase, this absorbs all this.

I think we’re going to have a huge spending boom next year and I think for the first time, and I know this is a sharp minority view here, for the first time in over two decades, we’re going to see inflation.

This is what Siegel is referring to: the truly extraordinary increase in the money base as a result of the CARES act and the Fed’s balance sheet expansion all with freshly printed money:

To put the current monetary thrust into perspective, M2 rose some $500B in the 2011 QE episode and another $756B on average during the next 8 years. It is up $3.2T just in the last 7 months!

In layman’s terms, M1 includes money in circulation (cash) plus checking deposits in banks. M2 includes M1 plus savings deposits, money market securities, mutual funds, and other time deposits. M1 is used mainly for normal life transactions while M2-M1 lies in savings and investments.

Very simply, the theory is that if you put money into people’s pockets, they will spend it. If they spend so much and so quickly that the supply of goods and services cannot keep pace, demand exceeds supply and inflation rises.

But notice how the reasonably good synchronicity between M2 and real personal expenditures disappeared after 2000 along with the continued slowdown in real spending growth rates in spite of ever rising money supply:

If you don’t know much about MRP and the velocity of money, you likely don’t know Dr. Lacy Hunt who, from his perch at Hoisington Investment Management, has been a vocal bull on U.S. Treasuries for over 30 years. This is what he told Bloomberg on August 14, 2020:

(…) The pandemic will eventually go away, but the debt will remain. It’s been my view that over-indebtedness ebbs economic growth. Debt is a double-edged sword: It’s increasing current spending in exchange for a decline in future spending unless it generates an income stream to repay principal and interest.

We had four secular peaks in total debt to GDP. The 1870s, 1920s and 30s, 2008-09, and we’re going to set a new peak this year, which will take out the peak in 2008-09. The debt surge reflects both a rising debt and a decline in GDP. In the three earlier instances, the inflation rate fell very dramatically. We now have a new secular peak in debt to GDP occurring within 12 years of the prior secular peak, whereas before they were decades apart. That’s massively disinflationary. (…)

Dr. Hunt reminds us of Irving Fisher’s equation, (M2*V=GDP), which states that money times its turnover (velocity) is equal to nominal GDP.

A rather simple way to understand velocity is how many times a given amount of money injected in the economy will turnover, i.e. get passed along and used by various economic agents. When the government sends $1200 checks to every American adult, if they all leave it in their bank accounts, the velocity is nil and the payment has no real impact on the economy. But if the money is spent and the receiving merchant then spends on distribution and inventory replenishment, turnover rises and the economy gets rolling as the process repeats and broadens.

The  Federal Reserve can influence the monetary and credit aggregates but it has virtually no control over the velocity of money. If the Fed buys Treasuries from banks but banks see no point in lending, the Fed is pushing on a string. Money sent directly to consumers, directly into M2 as Milton Friedman said, carries much better odds of being spent and stimulate the economy.

Now watch how M2 velocity decelerated post-2000 after rising steadily since the 1960s and exploding in the 1990s. In 20 years, M2 velocity declined by about one third. In the first half of 2020, it dropped another 23% to 1.1, almost stall speed.

A sharp decline in M2 velocity means that the rescue money did not work its way into the real economy.

The steep drop in V in the spring quarter reflects the transitory fall in Treasury deposits but also the ongoing and far more significant decline in the marginal revenue product (MRP) of debt, the main fundamental determinant of velocity.

Total domestic nonfinancial debt, excluding off balance sheet liabilities such as leases and unfunded pension liabilities, surged to a record 259.7% of GDP in the first quarter of this year, 11.4 percentage points higher than the 2009 level when Lehman Brothers failed. Confirming economic research regarding diminishing returns of the overuse of debt, each dollar of debt generated only 38.5 cents of GDP in the first quarter of this year.

This result is defined as the marginal revenue product of debt (MRPD), which is down from 40 cents at the end of 2019. Each dollar of debt has generated only 13 cents of GDP growth for the past four quarters, less than one-half of the 26.5 cents generated during the final four quarters immediately before the recession that started in late 2008. (…)

We expect velocity to drop sharply in the second quarter then rebound in the second half of the year but not sufficiently to offset the fall in velocity in the first half. In 1934, Irving Fisher wrote that the velocity of money falls in heavily indebted economies. We believe that Fisher’s finding will be correct because his view is supported by the evidence and the rationale that the huge additional debt added this year will not generate an income stream to repay principal and interest. Accordingly, the reopening rebound in the economy underway will falter, leaving the economy with a huge output gap. Extreme indebtedness in the corporate sector is a micro-consideration that also supports this view. (Hoisington Q2’20)

Enough econometrics. The real world during the pandemic:

• During the lockdowns, people organized their home for remote work and in-home entertaining: they bought work software, electronics, subscribed to Netflix, bought Nintendo Switch consoles and games, etc.. What is the velocity of these purchases for the U.S. economy when software and game sellers can immediately meet infinite demand on line, book revenues with 90%+ margins, mostly received abroad, meaning that whatever money velocity there may be, it will happen abroad. As a case in point, U.S. imports of Personal, Cultural, and Recreational Services jumped 11.4% during the first 8 months of the year. Gartner reports that U.S. PC shipments rose 11% YoY in Q3, the first time in 10 years with double-digit growth. All foreign stuff with double digit growth while total U.S. consumer expenditures declined 3.5%.
• This cash adds to already large piles of cash in software companies’ accounts, awaiting to be used for buybacks and dividends to investors who will most likely reinvest in other financial instruments.
• Meanwhile, a large part of the rescue checks got saved, while governments green lighted people and companies to withhold payments for rents, credit cards and mortgages, preventing huge sums of money from churning across the economy.
• Finally, the Fed bought trillions of debt instruments with printed dollars that went primarily to institutional investors who also reinvested in other financial instruments.
• Oh, and the Fed is vowing to keep short and long rates to the floor until inflation returns sustainably well above 2%, unseen since the early 1990’s, totally dis-incentivising banks from lending and feed money velocity. Why take risk with little if any profit margins? Excess bank reserves may be less immediately potent than direct deposits, but when banks do lend, they create even more money lending their reserves multiple times.

(Bloomberg, Macrobond and Variant Perception)

In all, a large part of the rescue money is simply recycling itself in equity or fixed income markets or sits in precautionary savings. And a lot of what trickles into the real economy carries little or no multiplier. Pushing on golden strings!

In effect, M2 and bank deposits remain sky high, nearly 6 months after the rescue launch. We will see how that evolves as debt and rent moratoriums expire. Unlike after previous recessions, there is no big pent-up demand other than for some services, most of which will have to wait for the virus to go away.

However, unless the Fed reverses QE and starts selling securities (draining money supply), or the federal government raises taxes, the money bulge will hang around the economy. Is this, as Jeremy Siegel believes, significant unused buying power that will shortly be unleashed? Or will continued high unemployment, the unresolved pandemic, huge indebtedness and profound and lasting scars, result in much higher savings rates, very slow velocity and GDP growth?

Nobody can be certain but Lacy Hunt’s arguments based on econometric identities weigh more heavily on my scale than Siegel’s beliefs.

In mid-August, I introduced Jean-Guy Desjardins, CEO of Montreal-based Fiera Capital, a global investment firm managing $171 billion, as one of the top global asset mixers around. Jean-Guy’s top-down approach focuses on the economic cycle.

In brief, Fiera then saw a 65% probability of a medical solution to the pandemic by the end of 2020 or Q1’21. By then, the world will have developed a large 6% output gap that governments and central banks around the world will seek to close as quickly as possible through sustained stimulative fiscal and monetary policies, creating an “ideal environment of 3 to 5 years of highly visible and unchallenged global economic expansion” for equity investors.

The same output gap that Lacy Hunt says will remain after the economic rebound falters because of low velocity and excessive debt, Fiera sees as a clear incentive for world governments and central banks to aggressively seek to close. Fiera assumes governments will keep borrowing and central banks oblige and print as long as the output gap persists. On the other hand, Hunt asserts that stimulus policies will not work any more than the post-GFC experiments because of already excessive indebtedness.

So, our table is well set:

• Professor Siegel sees an economic boom with rising inflation;
• Fiera Capital sees an economic boom without problematic inflation;
• Hoisington Investment sees continued slow growth and minimal inflation;
• Rosenberg and Mauldin are sitting on their pile of Treasuries, totally scared.

Amazing debates between very smart people: two highly successful institutional investors and three well informed and experienced pundits!

Two weeks ago, Fiera Capital produced a Matrix of Expected Returns under 3 scenarios that would find takers around our table:

Interestingly, expected one-year returns on U.S. and international equities are below 8.0% (forex-adjusted) and pale against 30-40% collapses under the stagnation scenario which carries not insignificant 25% odds. Fiera’s probability-weighted returns for U.S. equities are -2.9% (-0.8% for bonds). Recall that the Rapid Recovery scenario (45% odds) critically assumes that “a therapeutic is discovered in the near-term and proves sufficient in gaining control over the proliferation of the virus.” Whether that deserves a 45% probability at this time would also likely get challenged at the table.

Fiera’s economic enthusiasm is thus tempered by equity valuations: their best scenario only provides a 7.9% return on U.S. equities and steers them towards the more cyclical Canadian equity market, highly dominated by Materials, Energy and Industrials which together represent 40% of the Canadian index and would clearly benefit from booming world economies.

Gerry, a reader in Singapore, wonders whether the Rule of 20 works effectively “when the Fed is backstopping the market? Do we not need to take the market distortion into consideration, as the Fed is a major market player and seems intent on remaining so?”

The Rule of 20 is not a timing tool but it provides an objective gauge of where equity valuations stand compared with history, essentially giving us an unbiased measure of valuation risk vs valuation potential reward.

For sake of complete objectivity, one should use trailing (real, actual) numbers to assess valuation. At the current 25.5 R20 P/E (index level/trailing EPS + core inflation), the S&P 500 is 27.5% above its “fair” (historical median) of 20. If the R20 P/E returns to its “20” median, which it always does, the valuation part of the equation will decline by 27.5%. This is the current valuation risk.

Obviously, if earnings are rising, valuation can more smoothly return to fair value. It can also occur if inflation is falling significantly although that generally means slow economic growth and potentially slow, even declining profits.

We don’t know when we will revisit the “20” fair value, but we know we will eventually. Will it be a 1987 crash or a prolonged speculative bubble like in 1998-2002? We are only dealing with odds here because we really do not know, and nobody really knows. For reference, since 1957, the time it took from a R20 P/E above 23 (15% overvalued) to the market peak was 16 months on average (range: 2-38), 12 excluding the 1998-2002 episode (range 2-30).

What about the Fed as a market distorter?

The Rule of 20 uses inflation as a proxy for interest rates (discount factor) to avoid real rate gyrations or manipulation. Inflation (core CPI) is not perfect but it is not manipulated and it is a fair proxy for interest rates.

Valuation is a discount factor incorporating interest rates but it is also a gauge of investor optimism about the future. If people believe that the Fed, or the government for that matter, has their back and will do whatever it takes to keep the economy rolling, that can keep P/E ratios elevated.

But the Fed is not omnipotent, can make mistakes and can be wrong. It pretty openly had our backs since the GFC but the economy never really accelerated, inflation remained muted, profits declined in 2015-16 and valuations fluctuated.

Investors are also a fickle group that can quickly get worried about the Fed, the economy, interest rates, China, inflation, profit margins, taxation, yaddi, yaddi, yadda. Recall that between 2010 and 2019, we experienced 4 major market declines: -21.7% in 2011, -15.3% in 2015, -11.8% in 2018 and another -20.4% in 2018.

The Rule of 20 has gone through most everything since 1927 and yet, the 16-24 range around the 20 median remains intact.

Maybe ZIRF, zero interest rates forever, and FAIT, flexible average inflation targeting, will keep equity valuations high for a while. In FEARFUL FEARLESSNESS, looking at the last 60 years covering all kinds of economies and financial markets, I observed that major valuation excesses only corrected when a hawkish Fed took it on itself to end the party. It’s like if, at very elevated valuation levels, having climbed the wall of valuation worries, investors feel euphoric, fearless and push for an even higher summit.

Can earnings rise quickly enough and sufficiently to bring valuations back to fair value before equity markets do the revaluation job themselves?

As we enter the Q3’20 earnings season, analysts and strategists have become uncharacteristically optimistic that earnings will beat again and that managements will sound more optimistic about Q4 and 2021.

This is rather important given that trailing EPS will decline another $10-15 by the time 2020 earnings are wrapped up. By March 2021, trailing EPS will be $130-135, down 15-20% from 2019. As a result, at 3500, the S&P 500 is selling at 27 times March 2021 trailing EPS and 21x 2021 estimates. Defying the (G)odds.

This further decline in trailing EPS through March 2021 means that the R20 Fair Value [(20 minus inflation) x trailing EPS], where the index would be at a R20 P/E of 20, will decline another 8-10% from its current level of 2650. As the chart below shows, the correlation with the R20 Fair Value is tight (correlation of 98%).

Can markets correct meaningfully, given the liquidity out there, and an openly benevolent Fed sitting quietly on its hands by the dance floor watching people get totally wild? KKR’s Henry McVey recently smartly noted what few people did (my emphasis):

The Fed is now pre-committed on rates, but it has no such pre-commitment on QE or macro prudential regulation. In fact, Powell noted that the Fed seeks to promote its employment and inflation goals while warding off any risks to financial stability (i.e., asset bubbles) “that impede the stability of our goals.”

Put differently, the Fed’s new framework seems to be about mashing the accelerator on the economy via rates, while maintaining the option of tapping the brakes on markets via macro prudential regulation (bank oversight, stress tests, leveraged lending guidelines, etc.) and QE scaling. This approach is new, and we believe it represents growing caution about retail trading activity, particularly in high growth sectors like Technology.

What McVey is saying is that while Powell was guiding us to ZIRF, he also quietly retained the right to drain liquidity if the FOMC judged that the market behavior was a risk to their objectives.

About inflation, “reflationists” will use this two-scale chart to illustrate the relationship between trends in M2 and core CPI :

But if both series are on the same YoY change scale, the impact of M2 fluctuations on inflation is not so obvious, especially post the GFC and the Fed’s various QE programs that many pundits said would boost inflation.

The Fed has clearly decided to put much less weight on its inflation mandate, feeding the reflation thesis. Lacy Hunt does not agree and he’d better be right on his deflationary call amid the sea of liquidity seeking velocity. Otherwise, his assets under management, all U.S. Treasuries at Hoisington, would melt away as both inflation and long term rates would take off.

This really has the looks of a bet with only a binary outcome.

Roger, a fishing buddy of mine plays a poker game he calls Blind Lou. Five cards down, no peeking before the first round. How lucky do you feel?

He also has a variant he occasionally plays, mostly late in the night when the fridge is empty: Compulsory Blind Lou. Everybody in, blind. How lucky will you be?

At the table, our above five smart investors are all in, in their own way, bulls or bears, loudly playing the influencers, some, perhaps, talking their book.

But in truth, they are playing Blind Lou, trump or no trump ().

Nobody really knows.

So Gerry, yes, follow the Fed. But also follow M2, the savings rate and inflation. And this virus…

Nobody really knows.

There will be blood.

Luckily, it’s not Compulsory Blind Lou. And it’s wise to keep some chips off the table. To keep playing. Because it’s fun. For now…

I miss you Roger, and Bill, and Neil. And Alain, Jean, Brian, Richard, Réal. And salmon fishing!