Sources of wealth

Your sources of wealth vary quite dramatically depending on how wealthy you are.  At the top of the tree (top 1%), equity – both private and public – makes up a much bigger share of wealth than it does for everybody else (Exhibit 1).  After 35-40 years of a rising gap between rich and poor, the top 10% on aggregate are now almost twice as wealthy as the bottom 90%. Obviously, these numbers are American, and the gap between rich and poor is bigger over there than it is in most other developed countries, but the trend is the same everywhere.

The other thing I noted when I first looked at Exhibit 1 is how important retirement savings are to the middle classes and their wealth. If we assume that wealth will indeed mean revert at some point, and we assume that financial assets and property will probably take most of the hit when it happens (as is always the case), we can conclude that the middle classes will be more negatively affected by this trend than anyone else. This is bad news because the middle classes underwrite economic and political stability – not just in the US but all over the OECD.

Why is wealth-to-GDP long-term stable?

The Federal Reserve Bank started to collate data on US wealth in the aftermath of World War II, and quarterly statistics are released towards the end of the following quarter, i.e. the latest numbers published at this moment in time are from the 30th September 2020. For many, many years, US wealth-to-GDP traded in a range of 350-400% with a mean value of about 380%. Since the mid-1990s, though, wealth has grown significantly faster than GDP (Exhibit 2). Not surprisingly, the sharp rise in wealth has occurred at the same time as the gap between rich and poor has widened dramatically.

Over the past 25 years, US wealth-to-GDP has climbed from about 400% to almost 600%. The only major setback was caused by the Global Financial Crisis in 2007-08.  Not even the dotcom crash in 2000-01 and the associated recession had much of an impact. Today, about three-quarters of all US wealth is accounted for by financial assets – mostly equity holdings (including closely-held, typically family-owned, businesses), mutual funds and retirement savings. By far, the most important non-financial asset on the balance sheet is property.

As I said earlier, the long-term mean value of wealth-to-GDP varies modestly from country to country. Why is that? Think of wealth as the total amount of capital available and think of GDP as the total output in an economy. Wealth-to-GDP is thus a gauge of capital-to-output and measures how much capital that is required to produce one unit of output. In other words, it is a measure of capital efficiency, and all countries do not deploy their capital equally efficiently.

In economic growth theory, the capital-to-output ratio is long-term stable. In the US, it has traded in a range of 3.1-3.8 times, depending on which types of capital you include in the numerator. For example, earlier growth models did not include human capital, but they do now, leading to a larger numerator and thus a higher capital-to-output ratio more recently.

By applying some elements from economic growth theory, which is a very hairy topic (I hardly understand it myself!), you can prove that the capital-to-output ratio is always in the range of 3-4 times regardless of how developed the economy in question is, as long as it is an open market economy.

The follow-on from that is that the capital-to-output ratio should be unaffected by the invention of new technologies and, consequently, so should wealth-to-GDP. We therefore know that the steep rise in wealth-to-GDP of the last 25 years is unsustainable and can only be justified if the underlying theory – economic growth theory – can be proven wrong, and nobody has been able to do so yet.

Examples of different wealth regimes

If you take another look at Exhibit 2, it is clear that Americans have been through several wealth regimes in the past 74 years (and so have the rest of us – I just don’t have the data to prove it).  As you can see, US wealth-to-GDP troughed in 1951, following which it rose for the next ten years. After peaking in 1961, it went through another downturn which lasted almost 20 years.

Then began a rise that continues to this day, only interrupted by the dotcom crash in 2000-01 and the Global Financial Crisis in 2007-08. In the first 15 years or so, the annual rise was very modest but, since the mid-1990s, wealth-to-GDP has risen substantially almost every year. More recently, as a consequence of the Federal Reserve Bank’s reply to COVID-19, risk assets have been in extraordinarily high demand, which has resulted in an almost vertical rise in US wealth-to-GDP, as you can see if you look at the right-hand side of Exhibit 2.

The annual rise in wealth, from the early days in 1947 to now, is 6.75% to be precise. That is the rise in total wealth, and that obviously includes property and other non-financial assets which, on average, have risen more modestly than financial asset prices. If only financial assets were to be included, a more dramatic picture on regime changes would emerge (Exhibit 3).

As you can see, the difference in financial asset wealth growth from regime to regime is dramatic. As Horace (Woody) Brock – the author of the paper behind Exhibit 3 – once said to me, “my father retired in 1966 and died in 1981, i.e. his retirement years were almost exclusively in the second regime. Meanwhile, my brother retired in 1980 and died in 2000. Whose golden years do you think were better?”.

I have deliberately not updated Exhibit 3 beyond 2000, as we are probably still in this (third) regime. One could argue that 2001-2008 constituted another regime of near-zero financial asset wealth growth (which is correct), and that we started a new, fifth, regime in early 2009. That said, one could also argue that, despite a few bumps in the road which were all relatively short-lived, we have never left the high return regime which started in 1981.

Are US equities really that expensive?

According to MacroStrategy Partnership LLP, as of the 30th of September 2020, US wealth-to-GDP stands at 583%, some 50% above the long-term mean value. There can be no doubt that the action taken by the Federal Reserve Bank last year to provide ample liquidity in response to the outbreak of COVID-19 is a major part of the reason why the ratio is currently at those lofty levels.

As is evident when you look at Exhibit 4, flooding the economy with liquidity, as most central banks did last spring, and none more so than the Federal Reserve Bank, has had the effect of driving risk assets to new heights despite rather poor economic fundamentals. The rise in risk asset prices from the lows in the spring of 2020 is quite remarkable.

Consequently, US equity valuations have reached levels rarely seen before. As you can see in Exhibit 5a, the US CAPE ratio is now almost 35. Only during the dotcom bubble in the late 1990s were equities more expensive than they are now. I have preferred to use cyclically adjusted P/E ratios rather than standard P/E ratios, given the extraordinary circumstances in 2020. The same analysis conducted on the basis of traditional P/E ratios would have resulted in even higher valuation levels at present.

US equity valuations are now approaching levels that are no less than three standard deviations above the long-term mean value (Exhibit 5b). Based on Exhibits 5a-b, it is therefore tempting to conclude that US equities are indeed in a bubble, but things are not always that simple. Let me explain.

The robust performance of the S&P in 2020 was very much because of the extraordinary performance of Big Tech. The five FAAMGs were, between them, up almost 50% last year through the 9th of November and accounted for a massive share of the overall performance of the S&P 500 (Exhibit 6).

That would spell bubble left, right and centre if FAAMG earnings were nowhere to be seen, but that is not the case. In fact, 2020 was an exceptionally robust year for Big Tech earnings (Exhibit 7). As society closed down in response to COVID-19, the FAAMGs benefitted – some of them immensely so. Only a few days ago, I heard that Amazon’s UK revenues were up more than 50% last year.

Whether we are in for more closedowns in 2021 or not, society will continue to digitise, and the FAAMGs will continue to benefit. The pandemic has resulted in changing consumer habits, and the FAAMGS stand to benefit for many years to come. I therefore believe that the very robust performance of Big Tech in 2020 can be justified, although the same is not the case as far as some smaller – mostly unprofitable – tech companies are concerned (more on this below).

I have been active in financial markets since I graduated from university in 1984 so have been through a few bubbles over the years. The very first I was confronted with was the Japanese equity bubble which collapsed in 1989. The next bubble I ran into was the dotcom (Nasdaq) bubble which burst in 2000.

After that, we had the US credit bubble which led to the Global Financial Crisis in 2007-08, and the biggest bubble I have ever experienced is still with us. That is the bitcoin bubble which hasn’t fizzled out yet. See Exhibit 8 below for an account of the biggest bubbles over the last 400 years.

When you have ‘enjoyed’ the experience of a collapsing bubble a few times, you realise that there is a certain pattern to all bubbles. Obviously, they don’t all behave exactly the same way, but I have noticed certain common characteristics in the bubbles I have been through:

1. Retail investors account for a growing share of trading as the bubble grows.
2. Access to credit is almost always too easy.
3. Intrinsic value is being completely ignored by most investors.
4. Option activity skyrockets, particularly purchasing of small lots of call options.
5. Trading volumes are increasingly disproportionate to market caps.
6. Irrational behaviour drives prices, i.e. equities rise for the weirdest reasons.
7. The bull market is fuelled by misinterpreted information.
8. Well-meant warnings from officials are being ignored (take for example Greenspan’s warning about “irrational exuberance” in December 1996).
9. Insiders increasingly distance themselves, as they can is no longer see the logic behind the rise in prices.

I am not going to go through them all in detail now, but take another look at #6 and then zoom in on Exhibit 9 below, which paints an almost identical picture of the performance of Tesla and Bitcoin over the last five years.  The bull case for Tesla is about optimism – solid economic growth, electrification of all vehicles, globalisation, etc.  The bull case for Bitcoin, on the other hand, is about pessimism – overextended government balance sheets, the forthcoming implosion of fiat currencies, etc.  If investors were entirely rational, the two would never move hand in hand.

I do indeed see bubble signs when I look at some smaller Nasdaq stocks, but I do not yet see a systemic stock market bubble. Away from Big Tech, the S&P 500 has not performed much better than elsewhere and, as I said earlier, the sublime performance of Big Tech can be largely justified.

Most of the present bubble signs are in very small, non-profitable, (mostly tech) companies that retail investors, for some reason or another, have fallen in love with. Therefore, should the tech bubble collapse soon, I would expect the impact on Big Tech to be short-lived.

That said, the tandem-like ride in Tesla and bitcoin is highly irrational, and I would expect (at least) one of them to cave in at some stage. The joint bull market in those two is a very powerful sign of irrational exuberance, as Greenspan would have put it. You can hear more about bubbles if you dial in to our upcoming March 2021 webinar on the 2nd March at 3:00 pm UK time. It is open to all ARP+ subscribers, and you can register for the webinar here.

What could cause the current regime to end?

The fact that US wealth-to-GDP is now almost 600% is a strong indication that we are in very choppy waters, but it says absolutely nothing about what will cause the ship to sink and when it will happen. What comes next is therefore pure speculation on my behalf.

There can be no doubt that US monetary policy has had a massive impact on the current wealth regime. Instead of letting financial markets reset whatever needed to be reset after the Global Financial Crisis, the Federal Reserve Bank has provided plenty of support to risk assets in recent years.

It all started in the 1990s with the (so-called) Greenspan put. Under Greenspan’s stewardship, the Federal Reserve Bank’s responded to financial crises by holding a hand under financial markets. Greenspan retired, but his put did not, and it continues to be used vividly to this very day. By far the biggest put ever exercised by the Federal Reserve Bank is QE, which has had a tremendously positive effect on all risk assets. It is therefore obvious that a change in monetary policy could spell the end of the current regime.

That said, even without a fundamental change in monetary policy, the current regime will probably end sooner or later. You can only push the envelope so far before you reach the edge of the table. Will higher inflation cause the end of this regime just like it ended the first regime in the 1960s? That is not as farfetched as many think, given the massive increase in money supply more recently.

There are virtually no examples of rapid money supply growth without it having a detrimental impact on inflation. So far, only asset price inflation has been affected, but simple logic would suggest that, once economic activities normalise and the output gap has been reduced sufficiently, consumer price inflation will also be affected.

The best counterargument I have come across so far has to do with the size of the Fed’s balance sheet which is still relatively modest. As you can see in Exhibit 10, when measured as a % of GDP, the US Federal Reserve Bank’s balance sheet is not out of line with other major OECD central banks’ balance sheets. In fact, it is on the low side.

One could therefore argue that, despite massive amounts of US money printing more recently, one should probably worry more about the inflationary implications of excessive money printing in Europe and Japan. The only problem with that argument is that (i) the output gap is much lower there than it is in the US, and (ii) risk assets have not risen nearly as much in those countries.

As far as the output gap is concerned, and with Biden keen to spend almost $2Tn to boost US economic activity in 2021-22, concerns are rising that he will actually overspend, which will cause the output gap to quickly vanish and inflation to become a problem again (Exhibit 11). On that note, as I have said before, I don’t think central bankers would mind for inflation to become a little bit of a problem, as it is a very convenient way to destroy some of all the debt we are saddled with.

There are indeed some scattered signs of bubble behaviour in US financial markets at present, but those signs are not yet at a systemic level, i.e. not at a level we should be overly concerned about. Furthermore, bubbles without much leverage will never cause as much damage to financial stability as leverage-infused bubbles will.

At this point, I should stress that, whether we are in a bubble or not, wealth-to-GDP must mean revert. It is economic law and entirely independent of bubble activities. Having said that, bubbles always raise wealth-to-GDP temporarily, only for the ratio to come down again when the bubble bursts. Therefore, it could affect the timing of the mean reversion process, whether we are in a bubble or not.

US GDP was $21.4Tn in 2019. Due to COVID-19, it was lower in 2020, but we won’t know until the 25th February exactly how much lower. For argument’s sake, let’s say that we land on $21Tn. 380% of that is about $80Tn, i.e. that is what US wealth must drop to complete the mean reversion process and fulfil economic law.

As at the 30th September last year, US wealth was $116.5Tn. That is the number that must drop to $80Tn (-31%). Given how indebted most governments are, I would expect them to fight tooth and nail to keep interest rates at modest levels, i.e. the most likely asset classes to take the biggest hit are equities and property. This is why we are not very active in property markets and why we keep a relatively low equity beta in our portfolios.

I will stop here but will urge you to read Aliber and Kindleberger’s book (see supporting literature next) if you want to do a deeper dive on this topic. Should you want to dig even deeper, you should look at Hyman Minsky’s work. Minsky, the distinguished American economist (1919-1996), was probably the father of bubbles and crisis models, so his work should be next on your list.

Niels C. Jensen

17 February 2021

Supporting literature:

Manias, Panics and Crashes – A History of Financial Crises, Robert Z. Aliber & Charles P. Kindleberger, 2015

The Disturbing Dynamics of US Net Worth – The True End Game of Investment Management, Woody Brock, Strategic Economic Decisions, 2006

The General Asset Allocation Problem – Incorporating the Implications of Wealth-Growth Regimes”, Woody Brock, Strategic Economic Decisions, 2008

Appendix

The Anatomy of a Typical Bubble

The following has been inspired by Robert Aliber & Charles Kindleberger’s book “Manias, Panics and Crashes” (2015 edition). This book is, to the best of my knowledge, the most comprehensive account of past bubbles and the reasons behind.

Some sort of exogenous shock to the system is usually required to form a bubble. That shock can be in the form of a new invention (e.g. bitcoin), or it can be the result of a displacement of some sort. In addition to the exogenous shock, easy access to credit is also required, so bubbles can usually only happen if banks are ‘complicit’.

I say “usually”, as the bitcoin bubble has gained strength more recently at a time where the global economy is anything but healthy and banks not particularly cooperative. How is that possible? You can find the answer to that conundrum in Exhibit A1 below. As you can see, US bank deposits exploded in early 2020 as a result of COVID-19. Households and businesses alike changed their spending patterns as a result of the pandemic, leading to a pile-up of deposits in banks. You can therefore argue that access to credit has been less relevant in this particular bubble, as many already had plenty of liquidity.

Let’s assume the system has just been hit by a big exogenous shock. Somebody has invented a fancy new device that has captured people’s attention all over the world. Let’s say for argument’s sake that we are talking about bitcoin. Businesses and households start to invest in bitcoin and, before long, they begin to draw on their credit lines to invest even more. This creates a positive feedback loop where the rise in bitcoin’s price creates more optimism, which will drive investors to borrow even more.

The underlying shock that starts the bubble can be pretty much anything. Bitcoin is only one example. Take for example the US equity bubble that led to the collapse on Wall Street in 1929. That was caused by the rapid increase in car manufacturing in the US in the 1920s. The Japanese bubble in the late 1980s, on the other hand, was caused by financial deregulation in Japan and repatriation of capital to Tokyo.

The bubble will grow and grow as euphoria escalates – often for much longer than you thought possible when you first spotted it – and the underlying drivers of the bubble often become more and more irrational. Let me give you one example of that: On the 27th of January, for reasons not entirely clear to me, Reddit users suddenly fell in love with all stocks on US exchanges which had a ticker symbol starting with BB. Companies as diverse as Build-A-Bear Workshop (BBW), Bed, Bath and Beyond (BBBY) and BlackBerry (BB) all skyrocketed in a matter of hours (Exhibit A2).

Eventually, the bubble will burst – bubbles always do – but I have learned from experience that it often takes much longer for that to happen than simple logic would suggest. To me, the worst example of that was the Japanese real estate and equity bubble in the second half of the 1980s. I had graduated only a few years earlier so was relatively inexperienced, going into this bubble.

I still remember to this day how more experienced colleagues of mine were ridiculed, one even fired, for not taking enough risk in Japan when the flame was still burning. These guys could see the bubble from thousands of miles away and refused to participate, but that decision came at a high price. This experience has been in the back of my head ever since – bubbles can destroy your career in more than one way.

It can be as costly to ignore the bubble as it can be to participate in it. The trick is to participate in the earlier stages but to find the exit door in time. The approach that has worked best for me is to monitor the reasons given for justifying the rally. As the bubble grows in size, even respectable analysts often chip in with their own explanation why the continued rise in prices can be justified.

I have found that, in the latter stages of most bubbles, these explanations often change from one week to the next, i.e. Mr. Flipper (no names mentioned!) may justify the rally in tech stocks, using one explanation one week, only to change that explanation the following week. When that happens, we have entered the mania stage, and it is time to exit!