dshort's Comments

Siegel vs. Standard & Poor's

Sat, 28 Feb 2009 12:13:37 -0500

To SteveTN and pelican

Note that Shiller's P/E10 used the 10-year average of real (inflation-adjusted) earnings. The use of real numbers helps to normalize for the period of high inflation (and accompanying interest rates) leading up to the secular market bottom in 1982. Thus it is reasonable to compare today's P/E10 with the comparable ratios in both inflationary periods, such as the early 1980s, and deflationary periods, such as 1921 and 1932.

Siegel vs. Standard & Poor's

Fri, 27 Feb 2009 23:04:41 -0500

Excellent article. I prefer the valuation method of Yale Professor Robert Shiller over the Siegel's approach. Shiller smooths the P/E calculation by using the earnings average of the previous ten years as the divisor. With this method, the historic P/E10 average is 16.3. The February 2009 close saw a P/E10 of 13.

Does this make the market cheap? With the Shiller calculation, secular lows around the 1921, 1932, 1949, and 1982 bottoms have been in the single digit range. Here's a link with charts to illustrate: dshort.com/articles/20...

Tuesday Preview from Europe: Reeling in the Years

Tue, 24 Feb 2009 07:57:39 -0500

Hi Mole -- The "Four Bad Bears" chart is from my website: http:dshort.com -- I'm happy for anyone to reprint it in return for an acknowledgment link, as I mention on my website. CR always does so. I request the same courtesy from you. /Doug

The Value of Financial History

Tue, 06 Jan 2009 23:49:39 -0500

For a really long view of U.S. financial history, I keep an inflation-adjusted chart of the S&P Composite average monthly prices dating back to 1871:

dshort.com/charts/SP-C...

I find this a welcome alternative to the daily chatter in the popular financial press. If reversion to the mean (or, more properly, the long-term trend) has relevance, then the next few years may figure prominently in future books on financial history. For some additional context:

dshort.com/articles/re...

Long Term Fundamental Value of Stocks Smoother Than Prices

Mon, 01 Dec 2008 18:27:50 -0500

Thanks, John. You get a similar perspective by simply drawing a regression through the data:

dshort.com/charts/SP-C...

Over long time frames, regression to the mean is inevitable. Fortunately, on an inflation-adjusted basis, the overpricing of the U.S. market is less grim than the nominal price would suggest:

dshort.com/charts/SP-C...

Still, regression to the mean after wide variance usually means overshooting on the other side.

Cheers,
Doug

How Impossible Is Market Timing?

Sat, 22 Nov 2008 17:15:51 -0500

Mebane, thanks for the instructive chart. Let me underscore the message: "This has been an instructive year to showcase the benefits of a market timing solution, namely, RISK MANAGEMENT AND AVOIDING LARGE LOSSES" [emphasis added].

Over the long haul (decades), buy/hold versus a monthly MA strategy is pretty much a wash. But If you compare betas, there's a big difference in favor of timing. I wish you had included exhibit 14 from your Quant "article" and I urge everyone to seek it out. This chart beautifully illustrates the portfolio risk management of following a monthly MA timing signal.

Of course no single strategy works best all the time. If one did, then everyone would jump aboard and then it wouldn't work at all. Theoretically, during secular periods of strongly trending price and major turning points (e.g., 1995-2008, 1927-1932, etc.), monthly timing will probably be superior to buy/hold:

dshort.com/charts/SP50...

During periods of range-bound sideways movement, it probably won't.

Over the last decade, aging boomers heavily weighted toward a buy/hold strategy have been financially devastated. For many the vision of retirement has been sadly altered.

Here's a little mental exercise for alpha seekers: Ted saved and invested for 30 years and was within 5 years of reaching his retirement nest egg target. Unfortunately, the year is 1928. A few decades later his grandson Ned saved and invested for 30 years and was about five years from his perfect nest egg. The year is 1982.

Who would have been better served by timing? Ted or Ned? Who would have been better served by buy/hold? Ted or Ned?

Now reverse the question: Who would have suffered more in each scenario? Would the chance of superior returns outweigh the risk of financial destruction?

The value of the monthly 10MA approach is not to get rich the quickest but to get rich with sharply reduced risk of financial tragedy.

Cheers,
Doug

Pierced Hopes for Stock Bottoms

Thu, 20 Nov 2008 23:11:26 -0500

"Wednesday was a rotten market day, in a rotten market month, in a rotten market year, in a troubled economy."

I couldn't agree more, and Thursday was even worse. The S&P has now snatched the gold medal from the 2000-2002 bear market as the worst decline in the history of the S&P 500 (which took its present form on March 4, 1957). The question now is whether it will emulate the grueling decline in the Dow in 1929-32:

dshort.com/charts/bear...

Frankly, my favorite position now is cash. When the S&P 500 rises above its 10-month moving average, then I'm ready to wade back into equities:

dshort.com/charts/SP50...

Meanwhile, capital preservation has trumped my quest for alpha.

Calling a Depression

Thu, 20 Nov 2008 22:46:22 -0500

Depression? Today's decline moved the S&P closer to the tipping point than decline since the birth of the S&P 500. Here's an disturbing comparison of the current S&P 500 with two post WWII bears and the Dow destruction of 1929-32:

dshort.com/charts/bear...

Hopefully we're nearing a recovery like we did after the massacres of 1973-74 and 2000-2002. But so far, today's market bears an eerie similarity to debacle that marked the end of the roaring twenties.

The thing that worries me most is the financial impact to boomers nearing retirement. The demographic implications are enormous. Delaying retirement may be difficult if a severe and extended recession sends the financially ill-prepared to the unemployment lines.

Deflation Watch, Day 2: Consumer Prices

Wed, 19 Nov 2008 15:17:39 -0500

The Department of Labor website may only publish data since 1947, but the Bureau of Labor Statistics began tracking the CPI in 1913.

The October CPI marked the largest monthly decline since January 1938, during the Great Depression. Since that time, only 33 months have registered a month-on-month decline of 1% or more. Thirteen occurred during the decade of the 1930s, 15 during the 1920s, and 4 between 1913 and 1919. The declines during the '20s and '30s coincided with periods of deflation, as this chart illustrates:

dshort.com/inflation/i...

This one-month reversal is dramatic, but only time will tell if it's an outlier or part of a trend. Here are complete sets CPI and inflation stats since 1913:

dshort.com/inflation/c...
dshort.com/inflation/i...

Cheers,
Doug

John Hussman: The Market Is Not in Uncharted Territory

Mon, 17 Nov 2008 18:18:27 -0500

If today's market rolls over into another Great Depression (which I doubt), I a more long-term strategy is excellent for portfolio risk management. I follow a 10-month moving average timing signal on the monthly close. Here's a daily chart of the Dow from October 1928 to January 1935:

dshort.com/charts/Dow-...

Here's a monthly chart over the same timeframe with some 10MA signals:

dshort.com/charts/Dow-...

This monthly MA strategy is not suitable for long-term sideways markets (and it certainly won't be of interest to a mutual fund manager). But it has worked reasonably well since the mid 1990s. It would have saved a lot of pain in 1929-1932, and it would have moved your to fixed income in November of last year.

Consumers Buy Into Disinflation

Fri, 14 Nov 2008 18:57:03 -0500

For a visual perspective on inflation and the historic periods of deflation in the US, here's a chart that I update monthly with the release of the Consumer Price Index (CPI) by the Bureau of Labor Statistics:

dshort.com/inflation/i...

The chart also shows post 1982 inflation if the BLS had the same method as it employed from 1913 to 1982. See shadowstats.com for a full explanation of the alternate CPI, which is tracked at that website.

Thoughts on Market Volatility

Fri, 14 Nov 2008 09:30:23 -0500

For a long-term perspective, check out the intraday volatility in the Dow since 1928. Over this 80-year period, the average swing is about 1.8%. There have been only 64 days when the intraday volatility exceeded 8%. That's right -- 64 days out of over 20,300 market days. If they were evenly spread, that would be about one 8% plus volatility day every 15 months.

Here's the amazing part -- fourteen of them have occurred since September 29th. The Crash of 1929 had only eight. Another thirty followed during the ten-year Great Depression. Four were clustered around the Crash of 1987. Only two happened during the nasty 2000-2002 bear.

Now, guess how many of these volatile days ended with a gain versus a loss. Find the answer here: dshort.com/

Portfolio Expectations: What a Difference a Year Makes

Tue, 11 Nov 2008 18:30:54 -0500

Thanks, Richard. This article is another reason why you're in my watchlist.

The 81-year time frame really resonates with those of us who study market history. And when analyzing long-term performance, it's also important to factor in the impact of inflation, as these two charts illustrate:

dshort.com/charts/SP-C...
dshort.com/charts/SP-C...

Thanks again!

The Obama Bottom

Thu, 06 Nov 2008 09:16:01 -0500

The S&P 500 hit its recent low eight trading days ago. If we look at the bear markets in the S&P 500 since 1950, we see that the bottoming process lasted anywhere from six weeks to eight months:

dshort.com/charts/bear...

Drawing trend lines on charts during bear markets is an entertaining pastime. But the impact of the looming global recession on today's markets may keep you busy redrawing those lines for many months to come. The real precedents could indeed date from the late 1920s.

The chart line I find the most troubling is a linear regression on a century or more of market closes. Take your pick:

dshort.com/charts/dow....
dshort.com/charts/SP50...

You can speculate about the new sheriff in town and past being behind us. But I think the past remains very relevant. Take a look at these two charts and ponder one simple concept: "regression to the mean," which frequently involves overshooting in the opposite direction.

Market Bottom: Historical Norms May Not Apply

Tue, 04 Nov 2008 17:15:47 -0500

"Has the stock market bottomed?" you ask. Good question, but you confuse the issue by raising the topic of historical recessions. The recession call by the NBER will be based on GDP and related statistics that have evolved in methodology over the decades since the organization was founded in 1920.

Consider, for example, that the eight-month recession during the 2001-2002 bear market ended in November 2001, according to the NBER, with the S&P 500 at 1,084.10. The index didn't bottom out until the following October at 776.76. That's a further decline of 28%.

A more productive approach to the question of when the current bear market will end is to examine previous bottoming processes. The key word is "process". Rather than a sudden single event, the bottom is better understood as a process over a period of weeks or months. An overview of the eight completed S&P 500 bear markets since 1950 reveals this bottoming process that lasted anywhere from six weeks to eight months:

dshort.com/charts/bear...

In four of these bear markets, the first low in the process was the actual bottom. In three, it was the final low. In the triple bottom of the most recently competed bear, the trough came in the middle. The process ranged in length from six weeks to eight months. There is no clear correlation between the depth of the decline and the length of the bottoming process. Likewise, the overall length of the bear market doesn't reliably predict the amount of time spent bouncing around the bottom.

Bears Have Rallies Too

Thu, 30 Oct 2008 09:51:17 -0400

Excellent. For a broader historical perspective on the relationship of intraday volatility to bull and bear markets, see this series of chart for the Dow from 1928 to the present:

dshort.com/charts/dow-...

The color coding follows the traditional definitions: A bull market is a 20% rally preceded by a 20% decline and vice versa.

Surviving the Short-Term to Participate in the Long-Term

Tue, 28 Oct 2008 20:35:21 -0400

I've been analyzing the Dow raw data from 1928 to the present, available from Yahoo! Finance. Here are some Dow charts I've not yet posted:

dshort.com/charts/dow-...

I wanted to understand the historical precedence for the amazing volatility we've recently experienced. You have to go back to the first half of the Great Depression to find volatility as dramatic as we've seen over the past few weeks.

After reading the dimwitted CNN Money article, I simply sorted the Dow percent gains since 1928 in an Excel spreadsheet. What I saw was utterly amazing. Here's a link to a table showing the 55 days since 1928:

dshort.com/tmf/dow-5-p...

Actually there were no 5% days in 1928. The first occurred a couple weeks before the Crash of 1929. There's a total of 54 days to date, all from the timeframe of '29 Crash and Great Depression except for one in 1970, two associated with the Crash of 1987, two near the bottom of the Tech Crash in 2002, and two huge days in the past few weeks.

Rather disturbing.

Surviving the Short-Term to Participate in the Long-Term

Tue, 28 Oct 2008 17:46:28 -0400

After I read your article, I saw the CNN Money lead after today's market close: "Dow's 2nd best day ever" the headline declared.

In the spirit of Black Swans, here's a sanity check on the CNN Money headline. Today's percentage gain is impressive -- the seventh best Dow return since 1928. But check out the dates of the top six:

1. 15.34% 3/15/1933 Great Depression
2. 14.87% 10/6/1931 Great Depression
3. 12.34% 10/30/1929 Two days after Black Tuesday, the Crash of 1929
4. 11.90% 6/22/1931 Great Depression
5. 11.36% 9/21/1932 Great Depression
6. 11.08% 10/13/2008 Two weeks ago
7. 10.88% 10/28/2008 <= Today

And it doesn't get any better after that. Number 8 was two days after the Crash of '87.In fact, the next 44 in order of descending gains were all during the Great Depression except for three -- two in 2002 during the Tech Crash and the first day after the Crash of '87.

Indeed, these are interesting times!

Why Today Is Different From the Inflationary 1970s

Fri, 13 Jun 2008 22:18:30 -0400

Yes, today is a bit different from the '70, but far less so than you claim. Here's a chart that shows the "official" CPI with the CPI if we were still using the same methodology in use during the stagflation era:

dshort.com/inflation/i...

We don't yet have wage inflation, which creates an inflationary spiral. But the level of consumer suffering is indeed approaching the '70s.

Just Your Average Bear Market

Thu, 12 Jun 2008 12:25:16 -0400

Using the S&P 500 as the gauge and a 20% decline as the benchmark, Here's a PDF slide show that demonstrates the wide range of duration, depth and contour of the eight bear markets since 1950:
dshort.com/docs/Bear-M...

Using the 20% decline benchmark, the S&P 500 remains above bear territory. Here's a table of stats on the eight bear markets since 1950:
dshort.com/articles/ne...

One additional stat not included in the table: The average length of time it took these eight bear markets to reach a 20% decline was 10 months following the previous high. At present the S&P 500 is about 8 months beyond the high set last October.

Time will tell, probably soon, whether the current market turns bear. If it does, we may have to wait a while before we see how it compares to these other eight.

Fed vs. U.S. Consumer: Will Inflation Go Higher?

Mon, 02 Jun 2008 14:08:55 -0400

"Will Inflation Go Higher?" asks James Hamilton.

If the Bureau of Labor Statistics were using historically consistent metrics, we'd already be well above the 7.8% rate of inflation (headline CPI) posted in October 1973 at the beginning of the Arab Oil Embargo:

dshort.com/charts/two-...

Capacity Utilization Figures Starting to Resemble a Recession

Mon, 19 May 2008 05:34:58 -0400

Investor newbies and others with short memories may view the S&P 500 March 10 low as the market bottom, no recession. But anyone who remembers the 2000-2002 stock market is probably more skeptical:

dshort.com/charts/SP50...

As the chart shows, there can be some dramatic rallies during an extended downturn. Hopefully we won't see anything like the 49.2% decline during the last bear market.

Since 1950, recessionary S&P 500 market bottoms arrived, on average, about 14 months after the previous high. The March '08 low occurred 5.1 months after the October '07 high. That's mighty early for a bottom, although there is a precedent: The 1990 recession, which lasted 8 months, saw the market bottom out a mere 4 months after the preceding high.

On the other hand, two recessions -- '73-'75 and '81-'82 -- were accompanied by a 21 month peak-to-trough timeline. Here's an overview of recession stat since 1950:

dshort.com/docs/recess...

Given the complexities of today's economic conditions and the elapsed time since previous market highs, both extreme optimism and pessimism are probably premature.

Follow-up on CPI

Wed, 14 May 2008 09:35:34 -0400

Actually, if inflation were calculated using the same methods in place before 1982, the current rate is approaching 12%. The chart on the ShadowStats home page uses the BLS calculation methods from around 1990. For a look at inflation as calculated using the original method (pre-1982), see this ShadowStats chart (scroll down to the CPI Data Series):

www.shadowstats.com/al...

For the big picture of inflation going back to 1915, see this link:

dshort.com/inflation/i...

Frankly, I think the BLS was justified in some of the modifications to their inflation calculation algorithm. But the downside is that it obscures the degree to which today's economy is beginning to resemble the grim stagflation of the '70s and early '80s.

It's a Rally and It Feels So Good

Fri, 02 May 2008 10:57:53 -0400

"Feel-good rallies are a normal part of the bear market landscape. This is either a run-of-the-mill feel-good rally or I am wrong and this whole financial crisis/housing price deflation/bond market distortion will turn out to be nowhere near as important as many people thought."

I think you're right and share your perspective. Here's a comparison of a 2000-2002 bear rally with the current advance -- the S&P 500 with a VIX overlay:

dshort.com/charts/SP50...

Perhaps we saw a long-term bottom in March. However, given the current economic climate, I remain wary.

Still No Light at the End of the Housing Tunnel

Wed, 30 Apr 2008 05:22:34 -0400

"Home prices will likely fall throughout 2008. Before they can stop falling entirely, their declines have to slow down. Because the S&P data is almost two months old, there is hope this may already be happening."

And what, beyond wishful thinking, is your basis for hope? For the big picture in home prices (literally) see this snapshot from last September:

img253.imageshack.us/i...

And to put that rathter symetrical '80s boom-bust cycle into a longer perspective, see this:

drewdelahoussaye.com/b...

In the '70s and '80s, the downtrend was fairly symmetrical with the uptrend. There's no reason the expect this bubble to behave any differently.

I predict a lot more tunnel before any light appears.

The Fed’s Dilemma: Rescue the Housing Market or Feed the Poor?

Tue, 29 Apr 2008 16:03:12 -0400

"Bail Out Homeowners (or) Feed the Poor" is a dilemma not likely to be acknowledged by the Fed. Have a look at the Fed's mission:

www.federalreserve.gov...

The buck stops at the (virtual) US borders, which doesn't include world hunger. In fact, hunger within the US probably wouldn't merit a Fed footnote.

I'm sure over cocktails and hors d' overs Bernanke would sigh sympathetically over the plight of the famished. But history will judge him based on the depth and duration of the current recession and forthcoming bear market.

Besides, "thin is in" -- right?