CIO Blog: 100 Years of Misery

“If you see a bear and it’s more than thirty metres away from you, you’ve no need to worry. The bear will run away…
If you see a bear, say, twenty metres away there’s still a good chance it won’t bother you. It’ll roar a bit just to let you know it’s there, but if you stand quite still it’ll probably get bored and go back into the forest….
Then, of course, if you turn a corner and see a bear and it’s five or ten
metres away..”
“Then, presumably, I should start to worry,”….
“You’ve no need to worry. Bears are the most predictable of animals. If it’s five metres away it’ll certainly kill you. There’s no point in worrying at all.”

From “The Roads To Sata” by Alan Booth

Bear markets are an inevitable part of a life in investment –  we  count  nineteen  of  them since 19191. We could, as investors, adopt the view of the lonely old woodsman in Hokkaido quoted above and treat them as random events from the natural world about which nothing can be done. Stop worrying and let the bear decide. After all, bear markets have always been miserable, yet have always been succeeded by long upswings where returns dwarf previous losses. On the other hand, since bear markets are man-made – some small semantic irony – we could study them for clues about what might precede them and take precautions. Indeed, this note is written on the assumption that increased worry on our part ought, in due course, to be rewarded by lower losses.

THE IMPACT OF MONETARY POLICY

Most bear markets have been preceded by important changes – usually in response to developments in the economy. Of the nineteen bear markets, no less than 13 were wholly or partly the consequence of changes in monetary policy (see table 1). Given record low interest rates and the enormous quantity of liquidity injections today, it is reasonable to suppose that policy moves by the US Federal Reserve
– the 900-pound gorilla of global capital markets – are likely to have major implications for markets.
Is the past a useful guide? Quantitative measures (1930,1933,1937,1946) appear to have a swifter impact than changes to interest rates. Perhaps this explains the tentative way in which this subject is discussed today. The more recent pattern of raising rates very slowly and in small steps (2007, 2018) echoes the pattern of earlier episodes (1929, 1957). The gap between the start of rate rises and the onset of the bear market in those times is long – between 20 and 37 months. More abrupt increases take effect much more quickly.
At the current time, it is overtly clear that the Fed wants to move gently and slowly, causing as little disturbance to markets as possible. Success is more than possible, but the risk that stubbornly high inflation, problems in debt markets or international developments demand a swift change in policy cannot be ignored and have serious consequences for markets.

Sources:
Bloomberg for Dow Jones Industrial average from 1919 to 1932 and S&P500 average from 1932 onwards. Federal Reserve of St Louis for Federal Reserve Discount rate from 1919 and for Federal Funds rate from 1930.
Notes
1.    Some count the whole experience from 1929 to 1932 as one episode, though that overlooks the near 50% gain from November 1929 to April 1930, and also removes the opportunity to look at policy after the very sharp fall in 1929.
2.    The Federal Funds rate was the main instrument of market focus, the Discount rate only moved from 10% to 11.5% in 3 moves over the same period.
3.    The S&P500 made a fractional new high in October 2007, by when the Fed had already eased a couple of times. The bear market can be dated to October but that ignores the sharp (12%) break in the index from July to August.
4.    The rate rises ended a year before the bear market began.

THE EXCEPTIONS

As noted earlier, six of the bear markets had nothing to do with monetary policy: three were wholly related to international developments; two were the (probably inevitable) consequence of the end of war; one appears to be self-induced (1919, 1939, 1946, 1962, 2011 and 2020). A further four were not preceded by increases in interest rates – though in three of those cases monetary policy was tightened in other ways. Given that monetary policymakers are so determined to not be the cause of a new crisis or bear market, it is all the more important to review the exceptions. What else can possibly trip us up?

1. 1919 to 1921.
A variety of causes led to this bear. They include: a disruption to patterns of demand following the end of the First World War; the return of large numbers of soldiers who went from paid employment to seeking employment – thus depressing demand and wages; falling agricultural prices as European production started to recover; and the effects of the Spanish Flu pandemic. The introduction of Prohibition in January 1920, which curtailed the legal activity of a medium sized industry, must surely have added to the downturn (and contributed to misery for some).
There was also a completely different attitude towards the economy and economic policy making. The Gold Standard mentality was still in force and there was a widespread assumption that the inflation of the war time years would inevitably be followed by deflation. The deflation of 1873       to 1896 had been followed by modest inflation in the years up to 1914. It was a generally agreed proposition that the reverse would now happen – and ought to happen; a 2% inflation target was still many decades into the future! In addition, the remedies applied to the economy were to balance the budget and raise interest rates. The Federal Reserve Bank of New York’s Discount Rate was raised from 4.75% to 5% in December 1919, i.e. after the bear market and economic contraction had started, and it continued rising to 7% in June 1920.
It is impossible to imagine today’s policymakers raising rates and cutting the deficit to deal with a depression.
The only modern parallel is the impact of pandemics. The Spanish flu is generally reckoned to have started in February 1918, ending in April 1920. It exhibited three waves in a pattern with which we have all become too familiar.
2. 1930 to 1932
The Discount Rate fell from 6% in September 1929 to 4% in March 1930. The Federal Funds rate, which also touched 6% in September 1929, fell below 3% for a brief period in November before rising above 4% over the turn of the year and settling around 3% in March.
Rates should probably have been cut further and faster. The NY Fed repeatedly pressed for the system to buy government bonds on the open market (i.e. Quantitative Easing) and were repeatedly rebuffed2.
Events now look very different. Then, rates were kept too high for too long and liquidity injections were insufficient. Now, many argue that rates have been (and will be) kept too low for too long.
3. 1933 to 1935
The Discount Rate and Federal Funds rates jumped higher in March 1933. This was not a consequence of interest rate policy. Rather it reflected an executive order by President Roosevelt, who had been in office for 36 hours, imposing a weeklong bank holiday during which it was not possible to withdraw, deposit or transfer money. The measure was put in place as a response to renewed weakness in the banking industry and the need to shore up the Federal Reserve itself3.
Fascinating though the episode is, it is hard to draw any parallel today. The economy was already weakening; rates were at rock bottom and the chosen remedies of economic support and public works had not been enacted, let alone taken effect.
4. 1937 to 1938
The discount rate had been unchanged since 1934, but the Federal Reserve doubled bank reserve requirement ratios in 1936 to prevent an “injurious credit expansion”. In addition, the Treasury sterilised gold inflows to stop them expanding the money supply4. Furthermore, the introduction  of the social security payroll tax and increases in personal taxation meant that there was fiscal tightening along with quantitative monetary tightening. Not surprisingly, inappropriately tight monetary and fiscal policy led to a sharp contraction and bear market5.
Policy today is thankfully the complete opposite of what it was then. If such measures were put in place today, a sharp contraction in the economy and a bear market would no doubt be the result.
5. 1939 to 1942
The Discount Rate was unchanged at 1% from 1937 to 1948 and the Federal Funds rate stood at 0.25% from 1937 to 1945.
The S&P drifted lower by 9% from October 1939 to early May 1940 and then fell a further 25% between the 9th and 21st of May as the scale of the German success in France sunk in. The bad news kept coming. The bear market finally came to a halt just before the Battle of the Coral Sea in early May 1942; after Midway in June 1942 the bull market never looked back.
The only parallel that I can make with today is the quality of leadership in the West in the 1930s. Better quality leadership in Britain and France, in particular, from the mid-1930s might have averted the war.
6. 1946 to 1949
There was no change in the discount rate, though the Federal Funds rate did move somewhat higher in an erratic fashion. The main monetary change was that margin requirements were moved sharply higher in January 1946. There was also a widespread feeling that the Second World War would, like its predecessor, be followed by a depression. In the event, there was no downturn, but there was a bear market6.
7. 1962
Despite no increase in official rates, the interest rate offered on the newly created negotiable certificates of deposit rose from 3 to 4% in late 1961, drawing liquidity away from the market. In addition, the expensive valuation of the market sparked nervousness, which fed on itself as the leading indicators began to fall in a way which suggested a recession was in the offing. In the event, there was no recession and the bear market was relatively short lived.
8. 2011
There was no change in interest rate policy, though the Federal Reserve had been winding back some of the liquidity injection measures which had been put in place in response to the near collapse of the financial system in 2008. The most important factor, however, was the effect of the Euro area sovereign debt crisis, which had been smouldering since 2009, and caught fire in the summer of 2011.
9. 2020
The pandemic took everyone by surprise. In our view it was not the pandemic, which had precedents, but the extent of the lockdown - unprecedented - which drove the market lower.

DOES THE VALUATION OF THE EQUITY MARKET MATTER?

One of our biggest concerns at present is valuations, which are expensive. However, while that is a red flag, it certainly is not a divine message. Bear markets have started when the market has been expensive, and they have started when the market has been cheap. The 1919 episode started when the P/E  ratio had just climbed over 107. The subsequent falls were severe, reflecting the extent of    a decline in output and prices of 9% and 11% respectively8. The 1980 bear market started at an even cheaper P/E valuation of 9. Perhaps that cushioned the extent of the fall which the very severe monetary tightening might otherwise have caused. Or perhaps the market was looking forward to the consequences of the pro-growth policies of the newly elected Reagan administration.
The two bear markets at the start of this century started from very different valuations. In March 2000, the S&P stood on a P/E of 28 but the subsequent decline was slightly smaller than the bear market of 2007, which started when the P/E valuation was 18.
It is hard to detect a consistent relationship. Nonetheless, there is a trend of bear markets often beginning in times of heady valuations, but it is far from an exact relationship. What is clearly true is that the depth of the subsequent recession and the problems uncovered in it do have a relationship with the severity of the bear market.

THE YIELD CURVE

If nothing else works, then surely we should just exit when the yield curve tells us the bear is closing in? After all, yield curve inversions have bene excellent harbingers of uncomfortable bovine proximity and most bear markets were preceded by pronounced flattening and, often, inversion at some points of the yield curve9.
It is supposed to work as follows. Short interest rates are very heavily influenced (if not actually controlled) by the central bank. The yield on longer dated bonds, although influenced by the level of short interest rates, is a freely set price which reflects the judgement of the bond market about future growth and inflation. When long bond yields are higher than short interest rates, the curve is said to be upward-sloping and – depending on the steepness of the slope – anticipating growth. When long bond yields are lower than short interest rates, the curve is said to be inverted and indicating a slowdown in growth and/or inflation.

However, the yield curve did not flatten or invert before the following bear markets:
•    1919: The market clearly anticipated that interest rates would be moved higher.
•    1933/1937/1939/1946: Perhaps these can be laid at the door of the great depression and the war.
•    1987: The curve was steepening as bond yields were rising faster than the Fed Funds rate.
•    2011: The curve remained steeply upward-sloping.
•    2020: Oddly enough, the curve inverted in 2019 as the market perceived that the Fed had much more easing to do as both growth and inflation remained underwhelming. This should not be interpreted as the market predicting the pandemic six months ahead of the first cases of Coronavirus!

WHERE ARE WE NOW?

Bear markets can come out of nowhere but the normal precursors – higher rates, quantitative restrictions on lending, tax rises - are not yet present. The strength of the recovery, the rapid growth of corporate profitability, the abundance of new technologies with their promise of new products and productivity all provide good support for equity markets, as does the rather downbeat mood amongst investors. But the expensiveness of markets, the threat of inflation, the unprecedented situations in public finances, international relations, and the pattern of work carry with them the risk of unforeseen consequences emerging suddenly.
There are reasons for hope and caution. We remain nervously optimistic, and think misery is likely to be postponed a while longer, particularly as policy remains generally supportive.

• 1 - Bear markets are counted in different ways. This note defines such a market as a 20% decline from peak to trough. For this reason, the 1976/8 decline, amongst others, pass without comment, but the 2018 bear is referenced. On this basis, we count nineteen bear markets since 1919.
• 2 - Source: Federal Reserve History
• 3 - Source: Federal Reserve History
• 4 - Gold was flowing into the United States. Without sterilisation, the gold inflows would have increased the monetary base which would have helped finance a growing economy.
• 5 - Source: Federal Reserve History
• 6 - Source: Federal Reserve History
• 7 - Source: Robert Shiller, Irrational Exuberance for all P/E data
• 8 - Source: James Grant, The Forgotten Depression
• 9 - Sources: US Treasury for bond yields and Federal Reserve Bank of St Louis for short rates in this section.

Download : KH CIO Blog 100 Years of Misery(PDF, 1170 Ko)