How Bad Was It? With Paths Forward

Summary

• Last week's stock market sell-off was the 12th worst in a 90-year history, and the 4th worst post-World War II.
• This article looks at how markets have performed in the aftermath of similarly negative weeks.
• There are a wide range of outcomes, but on average, forward returns are not overly concerning.  As time horizons extend, markets recover.

For many investors, the coronavirus-induced sell-off in U.S. stocks was the second, third, or fourth worst week of their investing careers. Last week's -11.4% return for the S&P 500 (NYSEARCA:SPY) was the worst since October 2008 (-18.1%) at the outset of the Global Financial Crisis. Professional investors at least in their early 40s may have experienced the crash after 9/11, when the S&P 500 sold off 11.6%. Before those two memorable weeks, the previous worst week that may have been experienced by some readers was the October 1987 sell-off when the S&P 500 fell -12.2%, including a frightening 20% sell-off on Black Monday of that week. Any other investor that has experienced a worse week prior to those 3 weeks had to be active during the early parts of World War II!

It was a historically bad week for the market. Graphing weekly returns for the S&P 500 and its predecessor indices back to 1928, and last week's negative return is a visible outlier.

Including the Great Depression, last week was the 12th worst performance on record in a dataset stretching back a whopping 4,806 weeks. Over the full dataset, 99.8% of weeks have been better than last week.

That is the historical view. What investors should be asking is what have forward returns looked like from weeks this bad. In the table below, I have examined the 25 worst weeks for the S&P 500. To the right, I have also calculated the returns for the next week, next month, next quarter, next year, and forward 3 and 5-year periods. The goal is to answer the question: "After weeks as bad as last week, where does history suggest we go from here?"

The table above lists the 25 worst weeks on record in descending order. For the sake of readers, I will walk through a row on this table to illustrate the data contained within. The second worst week on record (row 2) was the week ending October 10th, 2008 early in the Financial Crisis meltdown. That week, the S&P 500 fell 18.14%. The next week, the stock market rebounded 4.61%. Over the next 4 weeks, stocks were up 3.74%, giving back some of the gains from that first-week bounceback. Moving forward 13 weeks, or roughly one quarter, and stocks were down slightly (-0.22%). One year (52 weeks) from that October 2008 weekly drawdown and the S&P 500 had returned +22.34%. Over the next 3 years, stocks would rally 37.34% (11.2% annualized). Over the next 5 years, stocks rallied 102.31% (15.1% annualized).

On average, stocks have tended to be about flat (-0.04%) in the week after a historically bad week. However, of the 11 weeks worse than last week, 9 were positive, excluding only the German incursion into France in 1940 and the early days of the infamous stock market crash of 1929. While returns the next week average to something close to zero, there has been tremendous volatility around that average. Annualizing the standard deviation of these 24 weeks gives a 55% figure, roughly three times normal return volatility.

Over the next month, from a historically bad week, stocks have tended to be off slightly (-0.85%). Of the 24 weeks with forward returns tabled, 12 were up over the next month and 12 were down. The range of outcomes is roughly double normal stock market volatility.

Over the next quarter, returns were slightly worse (-2.81% on average). However, if you exclude the 12 weeks in the Great Depression in the dataset, returns were actually slightly positive in the remainder of the weeks.

Over the next year, stocks were up slightly on average (+2.2%). Excluding the negative weeks occurring during the Great Depression and stocks were up (+7.6%), still below trend, but respectable. After the worst weeks on record, stocks are up one year later more often than not.

Over the next three years, stocks were up 9.58%, on average, and 20.24% ex-Depression. The only negative period outside of 1929-1937 was during the unwind of the tech bubble from 2000-2002.

Over the next five years from historically bad weeks, stocks were up 43.5% on average, and 51.35% ex-Depression. As the time horizon extends, returns have tended to normalize back towards the trend (roughly 10% annualized) and tended to be more likely to be positive. Time heals all portfolio wounds.

From this information, I believe that the post-9/11 period might be an interesting analogue for investors. In that scenario, the shock event was difficult to price by markets before the event, had a significant and negative near-term impact on travel, led to a short-term disruption in normal everyday life, and forced a difficult to quantify risk premia to be placed on markets in the aftermath.

Evoking 9/11 memories are painful, and I do not aim to be disrespectful to the loss of life in that event, which is at this point similar to the global loss of life from the coronavirus to date. The post-9/11 week saw stocks down 11.56%, quite similar to last week's 11.44% swoon. Over the next one week (+7.82%), one month (+11.27%), one quarter (+18.97%), stocks did quite well in the 9/11 aftermath. After a bit of a sell-off in mid-2002 related to accounting irregularities at Worldcom and continued post-tech bubble volatility, stocks were again in the green over the next 3 and 5-year periods, albeit below historical trends.

This data is useful for me in framing the recent move. By all accounts it was historic. Markets will be volatile over the next several weeks and months. Investors should re-assess their tolerance for risk. While momentum could lead to continued negative pressure, returns will ultimately normalize, providing appropriately positioned investors opportunity.

Analyst’s Disclosure: I am/we are long SPY. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Disclaimer My articles may contain statements and projections that are forward-looking in nature, and therefore inherently subject to numerous risks, uncertainties and assumptions. While my articles focus on generating long-term, risk-adjusted returns, investment decisions necessarily involve the risk of loss of principal. Individual investor circumstances vary significantly, and information gleaned from my articles should be applied to your own unique investment situation, objectives, risk tolerance and investment horizon.

Comments

[Tiki Bar Capital]

Excellent article. Another factor that suggests SPX will be meaningfuly higher a year from now is the vigorous intervention of central banks. The post-Financial Crisis Federal Reserve seems always to jump in when support is needed, as it did last Friday with Jerome Powell's 2:30 pm press release indicating willingness to backstop the markets.

The Fed doesn't always get the timing right. Aggressively hiking interest rates led to policy error in 2018, which culminated in a brief cyclical bear market. But the Fed's rapid about-face that December sent markets into a "v" recovery and a series of new record highs.

I'm mentally girding myself for a brief global recession due to coronavirus. That may lead to a "retest" of Friday's lows or even something worse. Alternatively, the market may "look through" the downturn in the real economy to the inevitable rebound in which economic activity lost in the first half of 2020 is picked up in the second half. In the latter case, a great many stock investors will be frustrated that they didn't take advantage of last week's volatility.

Myself included. I bought a lot of stocks last Wed.-Fri. But I still have most of my dry powder. It's hard to take the plunge 100%, even with the VIX hovering around 50.

[Market Map]

The forward 52 week returns, and even 3 year returns, may have more relevance in this case, than a forward 13 week measurement ( as some "corrective" activity can take as long as 6 months ). Most of the negative returns listed in the 52 week and 3 year returns columns fell within combined negative stock market / negative economic trend periods, as identified by the two tactical variables described in Part 2 and the Appendix ( Chart 1 "Great Depression" section ) of the E Book ""Improving Asset Accumulation: Tactical “Buy & Hold” with Exchange Traded Funds" *.
As the stock market trend turned negative on the close Friday **, the economic trend still remains positive. However, if the stock market trend remains negative into month end March or April, AND the economic trend turns negative **, then Table 1, Part 2 * provides us with a guide as to what forward returns have been produced by equity assets and duration assets returns over the past 70 years.
As any geopolitical or market disrupting event rears its head over the course of a primary stock market uptrend, the event must have deep enough impact towards turning the components of the economic index and trend negative. Otherwise, the equity market has typically recovered back into positive trend, most likely continuing to higher highs, with the decline being short lived.

Much of the time within secular primary market uptrends, the equity market can remain "irrational" ( climb higher and higher, with sharp price declines along the way ) longer than an investor can stay invested in it.
. . . .
* tinyurl.com/y6z8njmp ( Paste link into browser )
** tinyurl.com/y9rrzral

[Andrew Feazelle]

Can anybody with a FAST Graphs subscription tell me how overvalued SPY is right now?
I'm guessing $280 is fair value.

[fujilomi]

Thanks for the article. Whew it was pretty bad. And could get worse if fear and virus control measures damage US consumer confidence which is holding the market up, already big damage elsewhere like in China. Guess the FOMO and window dressing guys needed to get spanked, good thing last year was so great so only a flesh wound so far [but pretty deep!] for those long all or most of last year.

[Steve C....]

Ploutos, nice overview. I'll share my views of SPX as a systems trader. Looking at my monthly charts. From high to low the SPX dropped 538 points or 15.8%. On a log chart the bottom of the uptrend channel is just below 2700. On a non-log chart it is around 2600. There was triple negative divergence going into the high in SPX on a monthly chart. On a monthly candlestick pattern chart of SPX January closed with a bearish shooting star candlestick pattern. A prudent trader would therefore have placed a stop at the bottom of the pattern at 3214. This would have led to closing positions in SPX and avoided the damage to one's portfolio. Although February prices rose above the Jan highs, the stop would have served one well. Shooting star monthly patterns were also present in the DOW, $COMPQ, The IWM was weaker and did not take out its 2018 high, as the others did. The weekly charts of the DOW, SPX and IWM penetrated their lower bollinger bands and should bounce in the next few days. From an Elliot Wave perspective it is easy to see 5 complete waves into the high in SPX at 3393.52. A 38% correction of the move up from 2009 would be 2352. A 50% correction would be 2030. A 61.8% correction would be 1708.

It is unlikely that corrective action is complete. It is my opinion that the 2352 area will be tested this summer into a 4.5 year cycle low in SPX. In early 2025 a 9 year cycle low is due in SPX. Although it has not yet been confirmed, it is quite possible for the SPX to lose more than half of its gains from the lows in 2009 over the next 4 years. Since the fall of 2009, eight and thirteen month EMA's have successfully kept one on the long side this entire time. At this time, they still maintain a bullish configuration. And given how fast the initial decline occurred such indicators did not have any time to even squeeze closer to each other. We will see how things play out in the coming months. The indicators which go with these monthly moving averages are currently showing a similar pattern of negative divergence as they did toward the end of 2007, which is concerning.
Also of concern, is a point I mentioned in prior comments, that an overlay of the 89 week moving average of the advance-decline line at the all time high in SPX was sporting negative divergence going all the way back to 2017. This is a typical condition going into major tops.

On a daily SPX from an Elliot wave perspective, I think a wave 1 down low will occur in the area of 2790. And then a wave 2 up bounce will occur perhaps to around 3120 to 3150, backtesting the broken trend line from the December 2018 low. IMHO that would be a ideal place to become more conservative in one's investing and preserve capital by selling positions. If someone is retiring soon a corrective period similar to 2000 or 2008 would be harmful to one's retirement options. At a minimum I suspect the SPX will retest the 2350 area of the December 2018 low. Hope these technical comments are helpful to a few people. All the best to all.

[RenoGuy]

The only comparison on that list which was during the "era of central bank intervention" and also a rapid drop from peak pricing was #14, and the forward returns don't look so hot. Prior to the central bank era of massive intervention, the other two that were essentially falls from all time highs were #7, #8, and #17. Of those, only one had a good 3-5 year outlook. For my money, I'll start buying dips when stocks prices have been down for a long time, people are utterly depressed about their stock portfolio AND we get the kind of week we had last week. Currently, it's just too close to a recent market peak for me to believe this is a long term buying opportunity. (Although you can probably make nice coin if you're a good trader.)