Analyzing Market Troughs and Rebounds 9 comments
-
Font Size:
-
Print
- TweetThis
To better understand the behavior of the markets and analyst recommendations during periods of significant market corrections, we examined historic data around the last 3 major corrections (10/86 - 12/88, 7/89 - 10/91, and 9/99 - 10/03) and compared the data to the current environment. In addition, we analyzed the historical index data for the Dow Jones Industrials (from 1929) and the S&P 500 (from 1950) looking for periods when the indices declined by 25% or more from previous highs. Afterward, we gathered each index’s performance and volatility at the index’s peak, trough and 3, 6, and 12 month periods from each trough.
The graphs below contain the Percent of Analyst Revisions Up and Down (vs. the prior 4 week values), and the equal weighted cumulative return of all of the stocks followed by the analysts during the respective periods. The 1987 and 1990 downturns were relatively brief, while the 2000 – 2002 downturn was stretched over a longer period of time as it took several years before the effects from the bursting of the large cap and tech stock bubbles, 9/11, and accounting scandals worked their way to the small cap and non-tech universes.
click to enlarge
After examining the three graphs, two simple conclusions stand out. First market troughs are characterized not only by increases in downward analyst revisions, but also decreases in upward analyst revisions. Second, while markets tend to have sharp rebounds from their bottoms, it takes several months before a sustained rally takes place.
The next graph shows analyst revisions and the equal weighted market return for all stocks through the end of November 2008. The months from September – November 2008 have seen a radical drop in upward analyst revisions and an increase in analyst downward revisions. In fact, the spreads between the upward and downward revisions for October and November 2008 were greater than in any of the previous 3 corrections. This pessimism is indicative of a likely bottom forming as analysts estimates have finally started to reflect the poor economic conditions. Only after the expectations are driven far enough downward, will companies once again be able to surprise positively.
To get a better understanding of the data we broke the universe of stocks into four categories: Table 1: All, Table 2: Value/Growth (determined by the median P/E), Table 3: Size (where Large/Mid/Small = inflation adjusted market caps greater than $2 billion, between $2 billion and $300 million, and less than $300 million, respectively), and Table 4: Sectors (from Zacks).
The tables list data for each of the past 3 major corrections and the current correction separately, except for the Sector table which (for brevity) was aggregated across the 3 major corrections and presented alongside today’s values. The tables contain the net analyst revision changes (“Analyst Change”) and equal weighted holding period returns (“HPR”) for the periods from Peak to Trough, Trough to 6 Months, and Trough to 12 Months.
To smooth out the noise in the data, in general, the Analyst Change is averaged over 3 months and is calculated as follows: if at the Peak the upward and downward revisions were 30% and 40%, respectively, and at the Trough the upward and downward revisions were 15% and 60%, respectively; then the Analyst Change from Peak to Trough would be -35%, or (15%-30%) + (40%-60%).
Table 1 shows how difficult the current environment is. Through November the current correction’s Analyst Change and HPR has dropped more than in the preceding corrections. It has been particularly hard on Value Stocks (Table 2). As you look at Table 2, you should also note that Value stocks have been the clear winner coming out of the trough, even though they did not always underperform Growth stocks going into the trough. Table 3 illustrates that coming out of a correction, the Small Cap universe dominated the Large and Mid Cap universes. It has, however, generally underperformed its peers on the way down (except for the 9/99-10/03 period, when Small Caps had not participated equally in the prior market run up of Large Cap and Technology Stocks).
As mentioned above, Table 4 presents the Sector data as an aggregate of the prior 3 corrections versus the current correction. In addition to the Analyst Change and HPR data, we have presented the Ranked HPR (where the lowest rank had the highest HPR, and the highest rank had the lowest HPR) for each period and then sorted the Sectors by the Highest Average Rank (i.e. Lowest Average HPR) in the Peak to Trough period.
As you can see, the Sectors at the top and bottom of the Table generally make sense with Retail and Consumer Discretionary at the top, and Consumer Staples and Utilities at the bottom. The current data (with generally larger ranks at the top) also roughly coincides with the prior 3 corrections’ Peak to Trough Average Rank. By the time you reach 12 months beyond the Trough, however, the rankings are reversed with the beaten up Sectors leading the way out of the corrections.
Tables 5–8 examine prior periods in the Dow and S&P 500 where the market had a correction of 25% or more from its previous peak. The Dow data went back to 1929, while the S&P 500 went back to 1950. Here we examined the days from Peak to Trough, the HPR over different periods, and the volatility (determined by the Average Weekly Standard Deviation of the daily returns). Although it has seemed like forever, we currently are still on the lower end of the days from Peak to Trough. We are, however, on the upper end of the Peak to Trough HPR loss and volatility (although current volatility value may drop if the Trough moves further out from the highly volatile October and November, 2008, time periods). Finally there is some hope, in that 12 months after the Trough the indices had generally recovered a substantial portion of what they had lost.
Conclusion
The current correction is more severe than any correction since the Great Depression. In a relatively short period of time, the market has lost approximately 50% of its value while experiencing record volatility.
While each correction is different, they still exhibit some common characteristics/trends. Past corrections’ troughs can be marked by a significant increase in downward revisions, a decrease in upward revisions, and large increases in volatility. Coincident with the large current market sell-off (beginning in September and continuing through November 2008), analysts have also radically increased their negative earnings revisions and decreased their positive earnings revisions. In addition, in past corrections the sectors that suffered the largest HPR losses are Retailer and Consumer Discretionary stocks, while those having the least HPR losses are non-discretionary sectors like Utility and Consumer Staples stocks. The current correction has experienced similar losses.
In past corrections, the change in analyst revisions (particularly the decrease in upward revisions) and the excessive volatility have typically signaled a trough. This probably happens because expectations have to be driven down far enough for companies to start to surprise positively (which will most likely begin as “not as bad as forecast” announcements). Further confirming signs of coming out of a market bottom would be decreasing volatility and an increase in upward analyst revisions.
If this correction behaves similarly to prior corrections and has reached its trough, then over the next twelve months investors can expect to recover a significant portion of what was lost. Value should outperform Growth, Small should outperform Large, and the sectors that performed the worst going into the trough should be the better performing sectors coming out. Whether an investor chooses to reallocate within their portfolio now or prefers to wait for further confirming evidence of a market recovery, at a minimum the current environment offers him an opportunity to buy quality companies in these areas at a significant discount.
Related Articles
|
-
The Seeking Alpha 100
See allThe top 100 stock
market authors
selected for publication
Seeking Alpha 100 » -
Fastest Climbers
Authors with the largestSee all
increase in followers
in the last 7 days
Fastest Climbers » -
Top Commenters
Top commenters,See all
as ranked by their peers
Top Commenters » -
Top Instabloggers
The 100 most active
Instabloggers.-
1
TraderMark
-
2
Don Dion
-
3
Cliff Wachtel
-
4
Mike Havrilla
-
5
James Quinn
Top Instabloggers » -
1
-
Top StockTalkers
TopSee all
StockTalkers.
Top StockTalkers »
This article has 9 comments:
One area that deserves an additional observation is volatility. When they refer to record volatility, I assume they mean daily (and possibly intraday) volatilty. Looking at volatility from the point of view of 10% corrections, this bear market has had only three (so far), tying with the 1973-74 bear market for the fewest of any secular bear market since 1974.
Looking at volatility from the view point of 20% corrections, if the November bottom holds this bear market will also tie with the 1973-74 secular bear market for the fewest (none). If the bottom of this market comes with only one drop below the November 20 low, this market could end up with only one 20% correction (just barely for the S&P 500 and NASDAQ Composite, but none for the DJIA), still less than all other secular bear markets since 1900, except for 1973-74.
So daily volatility has been unusually high, but longer term measurements have been low. I will defer to an investor psychology expert to discuss what significance, if any, this has.
From an investor perspective, however, both economists and analysts are still substantially understanding the negativity in the economy and markets. We seem to have a long way to go to that trough and a possible economic or market rebound.
To extend Lilguy's comment, we're still very much in the learning phase of what is anything but a 'regular' cyclical downturn, with policy-makers in 'suck it and see' mode. I'm not convinced that arguments such as "If this correction behaves similarly to prior corrections and has reached its trough, then over the next twelve months investors can expect to recover a significant portion of what was lost" hold much water. I'm not saying the author's conclusion is wrong - just that I'd want to see an awful lot of confirmation to back-up initial conclusions based on entrails drawn from historic markets operating in entirely different economic circumstances, with entirely different market participants (and psychologies).
Just what is different is the great question and that is why you are seeing so much discussion, both logical and illogical. This is a great time for economic/financial/inv... junkies to be alive and seeking education. I just hope we can all afford the tuition.
Seeking Alpha does strange things at times to slashes and quotation marks.
On Dec 20 11:10 AM prudentinvestor wrote:
> Great analysis, but I really think this time its a bit different,
> as the combination of forces and events is dissimilar.
On Dec 21 02:52 PM BrucePile wrote:
> To say that this downturn can be modeled after the ones in the past
> is to say that debt doesn't matter. Nobody is saying that global
> debt isn't far and away more ponderous now than at anytime in history.
> I don't think you can say that we're not in the unwind phase of that.
> So how can we say this downturn must behave per past episodes? Not
> that we won't have bear market rallies that run for months or years
> (one of which may be starting now), but we are likely getting back
> to markets that are not artificially powered by debt expansion and
> that may take some doing.