The COVID-19 market correction has been on a scale not seen since the Great Recession. Using charts available on the Eikon coronavirus app, March’s Market Voice examines previous epidemics and the influence of current monetary conditions.
- The 11.5 percent one-week drop for the S&P 500 during the last week of February has only been exceeded on three occasions since 1960.
- COVID-19 seems to have more in common with the 1968 Hong Kong flu epidemic than the 2003 SARS outbreak.
- Most bear markets have taken place at times of tight monetary policy, whereas the current market correction is during a period of low interest rates.
The U.S. stock market seemed unconcerned about coronavirus (COVID-19) until the last week of February. As shown in Figure 1, the market rallied through the early January identification of the virus and the first reported death in China.
By mid-January, the rising death toll and Wuhan quarantine created concerns about a slowdown in global growth that was reflected in the sharp drop in copper prices to a near three-year low.
The S&P 500 (SPX) initially sold off in sympathy with copper, but the rally resumed in early February despite the World Health Organization (WHO) declaring COVID-19 a global emergency. The market made new highs even as the death toll exceeded that attributed to the SARS epidemic.
Figure 1: Coronavirus (COVID-19) timeline and market (lack of?) response
As February closed, the infection finally impacted the U.S. stock market, causing the biggest weekly SPX decline since the Great Recession. The stock market has since suffered further declines and volatility as the cases of COVID-19 continue to rise.
While the slump reflected concerns about the COVID-19 threat to global growth, it is hard to pin the sell-off to any specific virus-related event or announcement.
Instead, as is shown in Figure 2, the trigger for the COVID-19 market correction may have come from a subtle U.S. Federal Reserve tightening of monetary conditions.
Impact of Federal Reserve policy
The Fed overtly shifted to a more accommodative policy in the third quarter last year when the Fed funds rate was cut for the first time since 2008. The Fed further abetted monetary accommodation at the start of the fourth quarter by expanding its balance sheet.
The stock market responded positively to the Fed easing and the growth in the balance sheet sustained the market uptrend even after Fed funds bottomed in November.
The Fed halted balance sheet expansion at the end of the year and while the market made a new high in February it was on declining momentum. More importantly, both the January and February market peaks were linked to balance sheet contractions.
While the Fed was probably not directly responsible for the market decline, the (perhaps unintended) contraction in liquidity may have made the market more vulnerable to virus-related bad news.
Figure 2: Monetary conditions and the stock market rally (weekly closes)
The seeming good news is that if the decline was triggered by a modest move to tighten monetary conditions then the recent dramatic Fed shift back to easing should allow a resumption of the bull market trend.
While the market bounced in early March in anticipation of the Fed — and other central banks — easing, the benefit was short-lived. As noted above, the market has gone back into steep decline despite the efforts of the Fed to supply support and the Trump administration’s promises of fiscal action.
We posit that monetary tightness made the market vulnerable to bad news but was not itself the bad news. A resumed bull market would imply COVID-19 is not a significant factor for market performance; the continued doldrums for copper prices argues otherwise.
Watch: Coronavirus Correction Series — Episode 1 — An Introduction
Is COVID-19 playing out like SARS or the flu?
While COVID-19 and SARS (Severe Acute Respiratory Syndrome) are both coronaviruses and distinctly different from the flu virus, there are significant deviations in fatality and contagion rates.
SARS was quite lethal; in its active period from 1 March to 31 July, 2003 WHO estimates there were a total of 8,437 cases of which 813 died for a very high mortality rate of almost 10 percent (by comparison, the common flu has a mortality rate of about 0.1 percent).
SARS had a high death rate but it was not easily transmitted so was relatively easy to contain. WHO identified SARS as a global threat in mid-March 2003 but only four months later in July the SARS threat was declared over.
COVID-19 is highly contagious and spreading faster than SARS. Figure 3 from the Eikon Coronavirus App demonstrates that the identified cases and deaths from COVID-19 already exceed the entire impact of SARS, even though it is still in the early stages.
The mortality rate of COVID-19, while lower than SARS, is still not well determined. WHO-sourced data suggest a COVID-19 mortality rate of around three percent — still dramatically worse than the flu.
But because COVID-19 can occur asymptomatically, the number of actual cases is probably substantially higher than reported, so the actual mortality rate is probably lower.
According to the Chinese Center for Disease Control and Prevention, while the nationwide mortality rate is 2.9 percent the rate outside of Hubei (including Wuhan) is only 0.4 percent.
And in contrast to SARS, which was lethal for a broad spectrum of the population, COVID-19 deaths are largely patients above 65 with pre-existing health problems.
Figure 3A: A comparison of SARS and COVID-19 contagion
Figure 3B: A comparison of SARS and COVID-19 mortality rate
Even though it was not a coronavirus, the profile of the Hong Kong flu epidemic would seem to more closely match the experience to date for COVID-19.
In hindsight, the disease started spreading in Hong Kong and China in the first half of 1968, but it was not identified by WHO until mid-July. By late July it had spread to South-East Asia and well before the end of the year had evolved into a global epidemic — including the United States.
So like COVID-19 it was highly contagious and with a mortality rate close to the ex-Hubei China rate of 0.4 percent. Hong Kong deaths were also concentrated in elderly patients with health issues.
But because it was hard to contain, the case load was very high and roughly 18 months elapsed before the contagion faded — the overall death toll was more than one million people.
Figure 4 shows how the SPX performed in the context of these three epidemics. As with the death toll, the COVID-19 market correction has already exceeded any impact from SARS.
The stock market also initially made a new high in the early stages of the Hong Kong flu epidemic but then went into an extended decline, ultimately dropping roughly 17 percent from the peak. It then found support in the later stages of the virus.
Clearly, there were other drivers of the market than just the flu epidemic in 1968-69, nevertheless, this still should raise concerns about the poor potential for a resumption of the bull market while COVID-19 is still spreading.
Figure 4: Comparative SPX performance during COVID-19, SARS and HK flu epidemics
The COVID-19 market correction
The 11.5 percent decline in the last week of February was not just the worst since the Great Recession but there have only been three weeks since 1960 that saw a decline greater than this, and they make for rather frightening company: 10/10/08, 09/21/01 and 10/23/87.
These declines came during some of the worst market events of the past 50 years. In every case the market dropped another 20-25 percent after the extreme weekly decline before hitting bottom.
Before recommending to “sell everything” it is important to note some key differences between the recent weekly decline and its ugly brethren.
First in the three prior episodes the outsized weekly decline occurred well after the bear market was established, not at the start of the downtrend. More importantly, it is very rare for a sustained market correction to emerge when the Fed funds rate is not significantly positive in real terms.
The table below identifies all post-1960 “sustained corrections” which we define as the SPX being down at least 10 percent (on a month-close basis) year-over-year for two consecutive months.
This includes all events of traditional bear markets (down at least 20 percent from peak to trough) but also picks up some more modest but still extended declines (1965, 1976, and 1990).
In the Figure 5 table, the date of the “Market peak” is the highest month-end close prior to the identified decline, the “Total decline” is the drop to the ultimate trough, and “months” is the period from peak to trough.
Figure 5: Extended Market Corrections
|Market peak date||Total decline||Length in months||Real Fed Funds at peak|
“Real Fed Funds” is the nominal Fed funds rate less trailing annual CPI (excluding food and oil) inflation. It is apparent in Figure 5 that onsets of all bear markets and — except December 1976 — extended market corrections, emerged in an environment of tight monetary policy.
Fed funds were generally at least a percentage point above the rate of inflation and, since 1980 it has been at least two percentage points above inflation when an extended downmarket commenced.
Based on this history, it seems unlikely that we are at the beginning of an extended market decline. But there is the one exception of 1976 where the shock of the second oil crisis sent the stock market into its only prior sustained decline in the absence of a positive real Fed funds rate.
Like COVID-19, the second oil crisis created substantial dislocation in global supply chains and output, causing sharp slowing in GDP growth both globally and in the U.S. from 1974 into 1976.
Although the US economy started to rebound in 1978, the stock market continued to sink into early 1978 in conjunction with the rise in oil prices. The overall story of the market decline is more complex but still suggests that it will be difficult for the market to recover until the negative impact of COVID-19 on the global economy fades.
The fact that OECD has downgraded its 2020 global growth forecast by 50 basis points does not bode well for the market. But it is worth noting that the overall decline in the market was modest enough that it does not qualify for the traditional definition of a bear market.
Figure 6: Oil Prices and the SPX during the second OPEC crisis
Source: U.S. Energy Information Administration and Eikon
The bottom line: A modest potential downside?
The high degree of uncertainty over the outlook for COVID-19 makes it hard to have conviction on a bottom line.
To date, the COVID-19 market correction seems to have more in common with the 1968 Hong Kong flu epidemic than the 2003 experience with SARS. If so, it suggests that even with the Fed easing, the market is likely to struggle while COVID-19 continues to spread and the number of cases are increasing.
The end-February weekly decline was unusually large and historically only seen in the context of substantial market declines — greater than 30 percent — but the fact that it is emerging in an environment of low real interest rates points to a much more modest downside.
Both the market reaction to the Hong Kong flu and the shock of the second oil crisis in the mid-1970s suggests that the total downside from the February peak should be in a range of 15-20 percent, which would indicate that we are pretty near to the bottom.
The market is likely to remain volatile near-term and a resumption of the bull market will probably have to wait until the virus is clearly under control, which may not be until sometime in 2021.
How to track the market developments:
The Corona Virus app in Eikon is your single destination to keep track on the key market moving headlines as well as the charts, data and impact analysis on the markets, sectors and commodities asset classes. If you’re an Eikon user, simply search for ‘Corona Virus’. If you’re not a user, get access now or switch to Eikon.