Boom Times Compared

It has become abundantly clear by now that the late 2018 swoon was not yet the beginning of the end of the stock market bubble – at least not right away. While money supply growth continues to decelerate, the technical underpinnings of the rally from the late December low were actually quite strong – in particular, new highs in the cumulative NYSE A/D line indicate that it was broad-based.

Cumulative NYSE A/D line vs. SPX – normally the A/D line tends to deteriorate before the market peaks, as the advance narrows and fewer and fewer stocks participate in the rally. This did, in fact, happen shortly before the early October top.

To be sure, there are still technical warning signs as well – for instance, a number of divergences remain in evidence and several (but not all) measures of positioning and sentiment are back at extremes. On occasion of the two previous major market peaks in 2000 and 2007, the S&P 500 Index streaked to a marginal new high after an initial sharp correction and only began to decline in earnest thereafter. It is possible that such a retest of the highs will once again prove to be the turning point, but this is of course not certain.

What prompted us to write this post was the highly unusual timing and ferocity of the decline in late 2018. We were wondering whether there were any comparable historical precedents – and lo and behold, there is one. When the boom of the 1920s entered its final stretch, the stock market did something very similar: it fell quite rapidly and precipitously at the end of 1928 and recovered just as quickly. It then went on to make new highs in 1929. Alas, as is well known, this happy state of affairs did not last very long. It was rudely interrupted by a crash that was ultimately followed by the most severe bear market in history.

The sharp correction of late 1928 was just as unusual in terms of its timing as the one of late 2018 – and the market recovered just as quickly. The DJIA fell from a high of 295 points on November 28, 1928, to a low of 257 points on December 8 (down ~13%). By December 31 it was back above the starting point of the decline, ending the year at precisely 300 points.

Naturally, if the unusual December correction were the only parallel, it would not be worth mentioning. But there are several interesting parallels between the 1920s boom and today. Among them are valuations in terms of the Shiller P/E ratio or CAPE (cyclically adjusted P/E):

The stock market’s valuation is actually slightly above the level of 1929. This will be mitigated a bit when the 2009 data are no longer part of the calculation, but even then the market will remain close to the upper end of the long term valuation range. Moreover, today’s valuations happen to coincide with unusually high-profit margins, which are traditionally mean-reverting.

There are still more interesting parallels. For instance, in a recent missive, John Hussman quotes Dow Theorist William Peter Hamilton, who remarked on the stock market’s advance in 1929 that while it was clearly in the last inning, one had to respect trend uniformity in the short term, i.e., the fact that the Industrial and Transportation Averages were confirming each other and the advance was still quite broad-based.

Interest Rates and Central Bank Intervention

Even more interesting to our mind are the moves in interest rates and the actions undertaken by the Fed. Below is a chart of the NY Fed’s discount rate from 1915 – 1934 with annotations. You may be surprised to learn that the huge securities purchases after the 2008 debacle were actually not a completely untried policy.

The Fed engaged in large securities purchases from late 1921 to mid-1922 and again from mid-1923 to late 1924 and in mid-1927. This was done with the intention of stabilizing consumer prices  (obviously, this price stabilization policy continues to be popular with central banks to this day – and it is just as dangerous today as it was then). Interestingly, the rate hike cycle was subject to a pause between mid-1928 and mid-1929, which is also eerily reminiscent of the current situation. Lastly, the Fed once again engaged in massive securities purchases between the fourth quarter of 1929 and the first quarter of 1933 – only this time it was unable to keep prices from declining.

The next chart shows a weighted average of interest rates in New York from 1919 to 1933 (it contains the average rate on stock exchange call loan renewals, the  rate on 4-6 month prime commercial paper, the rate on 90 day prime bankers’ acceptances, and the rate on 90 day stock exchange time loans). It is noteworthy that market interest rates kept rising during the Fed’s rate hike pause. As usual, the Fed followed the market and was forced to “catch up” with it in mid-1929. By that time, the economy was already deteriorating.

Weighted average of NY interest rates from 1919 to 1933

The late stages of a boom are traditionally accompanied by rising interest rates. In essence, this represents a reassertion of time preferences, which are higher than previously thought. A distorted production structure ties up ever more final goods in production, and bottlenecks begin to develop. A scramble for scarce capital ensues, while still rising asset prices at the same time egg on speculative credit demand  (for additional color on this see also: The Capital Structure as a Mirror of the Bubble Era).

Eventually, the distortion of relative prices that is a hallmark of the boom reverses (although this never leads to a restoration of the same price structure that prevailed on the eve of the boom) and a bust commences. The bust is actually a salutary event – it serves to realign the economy’s capital structure to the actual balance between savings and consumption.

This realignment takes time and is accompanied by numerous unpleasant side effects, but if the economy is left to its own devices it usually doesn’t take overly long. Unfortunately, the administrations of Herbert Hoover and his successor FDR were both extremely interventionist – hence the “Great Depression” (this was not the main reason for the severity of the bust in the short term, but it did deepen and above all lengthen it unnecessarily).

What Happens Next?

The parallels discussed above do not necessarily mean that the pattern of the stock market will continue to track that of the late 1920s boom period. It is certainly a possibility one should not dismiss out of hand, but as mentioned above, the two most recent major market peaks occurred in the context of retest highs.

There are also differences between the various boom-bust cycles; for instance, the 1920s boom was characterized by both malinvestment in the higher stages of the production structure and debt-funded over-consumption on the part of consumers. This time around the consumer spending has been comparatively muted, while the corporate universe has amassed truly staggering amounts of debt and is home to an unprecedented number of “zombie companies” which depend on extremely low-interest rates (and careless investors) for their survival.

Obviously, the Fed’s back-pedaling in the wake of the October-December swoon in stock prices and the concomitant widening of credit spreads has emboldened investors. After all, they were told that “QT” will end early, that there will be a pause in rate hikes, and that the Fed stands ready to reactivate QE “if necessary”. There are even discussions over expanding the range of securities the Fed might buy to ward off the next bust.

However, investors would do well not to ignore history. As long-time readers of this blog know, we are of the opinion that the extent to which loose monetary policy can affect the economy and markets depends largely on the state of the economy’s pool of real funding. As long as this pool is growing – which permits the diversion of scarce resources into bubble activities through money printing – boom conditions can be restored fairly quickly. This is not the case once damage to the pool of real funding has become so grave that it stagnates or even shrinks.

This was essentially the situation the US economy found itself in when the market crashed in 1929. New York Fed President Benjamin Strong was one of the foremost proponents of the disastrous price stabilization policy (see also “The Errors and Dangers of the Price Stability Policy” for additional information on this topic). Moreover, he pushed for the 1927 discount rate cut and the associated securities purchases, which ended up catapulting the stock market bubble into overdrive (Strong famously quipped that he was providing a “coup de whiskey to the stock market”). He did so mainly to help Great Britain, which suffered gold outflows at the time.

Strong died in October 1928. Other Fed members were reportedly a lot more concerned about the speculative fever that had gripped Wall Street and it stands to reason that the decision to hike rates again in mid-1929 was at least partly driven by this concern. What happened thereafter should stand as a warning to investors who have become convinced that the Fed will always be able to bail them out.

Year-on-year rate of change in securities held by the Federal Reserve System from 1929 to 1933

As the chart above illustrates, the widely believed assertion that the Fed did nothing to counter the post-1929 bust is simply wrong (Ben Bernanke once even apologized for the Fed’s alleged inaction at the time!). Strong’s “coups de whiskey” became a veritable flood of whiskey – it was QE on steroids. Within three-quarters of the 1929 crash, the annual rate of change of securities purchases by the Fed reached almost 300%. A second wave of massive purchases began in early 1932 and by March 1933 the Fed’s securities portfolio had grown by a cumulative 404% from its level of October 1929.

And yet, the Fed was unable to keep the bust from unfolding. Admittedly, it has become easier for the Fed to boost the money supply in modern times (almost all commercial banks are members of the system nowadays and the FDIC helps prevent the disappearance of deposit money held at insolvent banks). We would contend though that in view of the effects the monetary pumping efforts of the 1920s exerted on the economy and markets and the failure of the same policies to gain any traction in the early 1930s mainly shows that the pool of real funding had indeed stopped growing by 1929.

Conclusion

The main takeaway from this is that QE is not always guaranteed to “work”. It stands to reason that after an extremely long time of (too) low-interest rates a lot of capital has been malinvested – and there are numerous indications that this is indeed the case. The danger that the economy’s pool of real funding is in trouble is accordingly elevated, in which case the Fed will be powerless to stop a major bust from unfolding once it begins (note that it is not possible to measure the pool of real savings – but one can certainly make educated guesses about its state).

Of course, a central bank in a fiat money system can in theory always opt for hyperinflation. Venezuela is a recent example of a country in which one of the worst economic contractions in history is accompanied by soaring nominal stock prices as the value of the local currency collapses (since we last wrote about Venezuela in September of last year, the IBC Index has risen by another 2,705% in bolivar terms  – a roughly 28-fold increase – while the economy has contracted by 25% y/y and the unemployment rate has soared to 45%).

The Fed as it is constituted today will certainly refrain from deliberately pursuing such a policy (although it may well trigger accelerating price inflation by mistake); in fact, it is illegal for the Fed to fund the Treasury directly – and this makes a big difference in terms of potential money supply inflation.

The future is unknowable and it cannot be ruled out that the central bank will one day be stripped of its independence if there is a severe economic downturn. That would increase the risk that a policy leading to hyperinflation could be adopted, but that is a bridge we will cross when (if ever) we get there. Until then a major bust should be expected to result in sizable declines in asset prices.

Charts by: StockCharts, acting-man, multpl.com, Frank Shostak; data by Dow Jones, St. Louis Fed, NBER