Skip to main content
Message Font: Serif | Sans-Serif
 
No. of Recommendations: 17
June 22, 1930 was a Sunday with no Journal to summarize, so once again I'm going to drop the mask of objectivity and make with a little editorial commentary and impressions of the past week.


Do the Country Shuffle - If there was a single item that made me sit up from my normally slumped position this week, it was the first one in the Market Commentary from Saturday:

Editorial noting that large US loans to foreign countries are beneficial to US business because the money gets recycled to buy US goods; this helps to absorb our production capacity, which is greater than domestic demand.

Now let's see, a country that has a big creditor position, a production capacity too big for its own people to absorb, and that therefore is loaning money to other countries to get recycled back into a trade surplus... might that remind you of any country today? (Hint: A Inch). Now, I'm ashamed to admit that until I saw that item it hadn't occurred to me that while the current situation might bear some resemblance to 1930, the countries might have to be shuffled around to really make things correspond - the flash of insight was blinding! Well, if China now is us then, then who then are we now? Let's see ... proud owner of the world's reserve currency franchise ... running sizable merchandise trade deficits ... I think I got it! (Hint: Baring Attire). OK, but then who now is Germany then? Let's see ... huge outstanding debts, postponing dealing with them as long as possible through a series of temporary gimmicks, government reaching a state of deadlock... got it! (Hint: Facial Iron)

Hmm ... that isn't exactly encouraging when it comes to what happens next ... although, on the bright side, at least in the case of California the consequences of having a strange looking Austrian in charge haven't been anywhere near as drastic this time ...


The Tariff Revisited - As expected, I got some adverse reaction to my opinion last week that the tariff didn't seem to have been a major factor in the decline so far. Let me clarify a couple of things. I'm not in favor of protectionism, nor do I think the tariff was a good idea - it probably wound up doing some damage. All I said was that in my personal, constitutionally protected opinion, it didn't seem to have been that much of a factor in the decline so far (I'm not excluding the possibility that it later did become more damaging). To review, I think this because:

- The trade crash already appeared to be well in progress before the tariff passed - in fact, this trade crash is even more remarkable given anticipation of the tariff, since we would normally expect a big bump in imports before the tariff was passed as importers stocked up as much as possible to avoid the raised duties.

- Tariffs were not exactly unknown at the time; many were already in effect around the world; as Mellon says this past week, previous tariff raises had also caused “gloomy prophecies”, but trade had nevertheless always increased in the past. It's also not the case that the tariff was hugely more punitive than previous ones - if I may cheat slightly and use an item from tomorrow, 80% of imports were duty-free or had duties reduced or unchanged under the tariff, and Hoover had made his intention clear to use the flexibility provision to address complaints about the remaining items.

- My final point was that the stock market declines seemed to be happening more in response to bad business news and commodity price declines than to the progress of the tariff bill. This is, of course, a subjective judgement and I invite the readers to come to their own conclusions.

Regarding this last point, however, I can't restrain myself from responding to one analysis that I was pointed to about the tariff's impact. This was by none other than the late supply side guru Jude Wanniski: http://polyconomics.com/ssu/ssu-010713.htm

Now, when I referred last week to a religious war among economists, I didn't expect to have to take on one of the Grand Ayatollah's himself, and without even a couple of warmup bouts with tomato cans. However, I will not shrink from my duty. Let me at least do a few pushups to warm up ... ninety nine ... a hundred ...

OK, here goes. With all due respect, this is one of the most bizarre and stupid analyses I've ever seen. A point-by-point refutation would be way too long, but here's a few of the high points. Mr. Wanniski starts right out with a whopper:

“The most common explanation of the crash is that the market was overpriced, the victim of heedless speculators who had somehow lost their grip on reality in the mad rush for quick profits. But that explanation has never quite satisfied, either empirically or logically. There is no real sense in which the market can be 'underpriced' or 'overpriced.'”

And here i've wasted all these years listening to suckers like Buffett or Soros who claim that the market has been known, on occasion, to go stark raving bonkers. Yes, it's all clear now. The market didn't crash because hot stocks like A&P were selling for 40-50 times earnings or more, and the most reliable measure of market valuation as a whole, the PE10, hit a peak over 30 not to be equalled until 2000 (see figure 1.3 in http://press.princeton.edu/chapters/s7922.pdf ). It didn't crash because traders were, if I may be excused for using a technical term, leveraged out the wazoo, with margin debt setting new records every week. It didn't crash because business started to decline. No, Mr. Market is always cool and rational:

“The market places a value on each company listed on its exchange, based on its calculations of that company's future income. ... If one accepts this rational model of stock market behavior, it's logical to believe that the market at its 1929 peak was exactly where it should have been, and that the crash resulted from some stupendous error ...”

So, what did cause the crash? What rational cause was there to reduce the valuation of all US industry by approximately 50% in around 3 weeks starting on Black Thursday, October 24? Clearly, the only explanation is:

“On Thursday, October 24, the anti-tariff forces suffer another setback; casein tariffs are raised 87%.”

Hmm ... casein. Well, it is one of the more commercially important milk-derived protein byproducts, but still, a 50% haircut for the value of the entire US industrial base does seem a little ... what's the word ... demented? To be fair, I guess Mr. Wanniski is arguing that it wasn't the casein tariff itself but that the market somehow at that moment realized that the anti-tariff forces were beaten and tariff passage was inevitable (although Smoot-Hawley didn't in fact pass for another nine months after much further sound and fury). But lets examine this a bit further. By Mr. Wanniski's theory, the mere realization by the market that tariffs were probably to be raised at some future time justified that 50% haircut. Now let's fast forward to Friday the 13th of June, 1930. Until that point, despite Mr. Wanniski's version of events, it appears that there still was some doubt about passage of the Smoot-Hawley tariff, since it still didn't have the votes to pass in the Senate (this is probably why there was an editorial against it roughly every other page in the WSJ). It's an item from that day that's really the final nail in the coffin:

Passage of the Smoot-Hawley tariff seems likely. With Senator David Reed, declaring his support, the bill is likely to pass in the Senate by two votes, and then to be adopted by the House. President Hoover is likely to sign in order to end “business uncertainty surrounding tariff consideration”; it's feared that a narrow defeat or veto would simply lead to the reintroduction of the issue at the next session of Congress. Senator Reed says the bill has been improved more than generally appreciated; his action followed meetings with Secretary Mellon and President Hoover.

This switch by Senator Reed and another Senator is what gave the tariff a majority - and, since he announced it after conferring with Hoover, it's clear that a veto is also not to be expected. So what was the market's response to this? If the mere probability of a tariff bill nine months later was enough for a 50% haircut, surely the inevitability of one in a few days would cause a cataclysm! Dow on Friday ... drumroll please ... up 2.51 points (1.0%).


I will skip the company review this week due to being worn out from wrestling with Mr. Wanniski, but I hope to have an expanded version next week. Until then ...
Print the post Back To Top
No. of Recommendations: 6
The market didn't crash because hot stocks like A&P were selling for 40-50 times earnings or more, and the most reliable measure of market valuation as a whole, the PE10, hit a peak over 30 not to be equalled until 2000

Any method that uses 10 year data from 1929 would overstate an average P/E because of the huge decline in earnings that begins post WWI and does not trough until 1922. From the trough in 1922, earnings growth is more than fivefold.

I'm also not sure why you declare such a metric to be 'the most reliable measure of market valuation'. From what I can see, such a measure moves rapidly between extremes, does not hover or settle near any one price point and spends varying lengths of time within 'undervalued/overvalued' territory. Especially beginning in the 20th century with the imposition and constant adjustments of the income tax and how such changes impact the value of debt vs. equity, dividends vs. retained earnings, real capital gains etc. Not to mention the near-pervasive inflation that was uncommon in previous measured periods.

It didn't crash because traders were, if I may be excused for using a technical term, leveraged out the wazoo, with margin debt setting new records every week.

Margin debt grew rapidly from 1926 through to Oct. 1929, roughly at the same rate of increase in the genereal market. If margin debt caused the bottom to fall out of the market, it could have did so in any single random month, but instead the market plunged the very same month that political leadership assured that the tariff would be passed, with each side blaming the other for not moving to pass the tariff bill fast enough.

Blaming margin debt is like putting the cart before the horse. Before stock positions are liquidated to satisfy margin requirements, the market is already declining as it is unlikely that margin investors would liquidate winning portfolios en masse. Margin debt can and probably does impact the extent of the decline but is not what precipitates the decline. Decades of stock margin data demonstrate that margin debt is a coincident or lagging indicator of the stock market, not a leading indicator.

If the mere probability of a tariff bill nine months later was enough for a 50% haircut, surely the inevitability of one in a few days would cause a cataclysm! Dow on Friday ... drumroll please ... up 2.51 points (1.0%).

Specific news that has been given the opportunity to be discounted rarely has a dramatic impact upon confirmation. It happens enough that an adage had been crafted to explain it, "Buy the rumor, sell the news!" -- or vice versa, in this case. It happens in all markets, not just stock markets. In handicapping sports wagering, the variables that could affect the outcome (injuries, field conditions etc.) are discounted upon recognition and reflected in the wagering line, instead of being adjusted solely on gameday.

For example, look at the health care stocks today. Back in September, these 'defensive' stocks sold off harder than the overall market, likely on recognition that the election would produce a political body that would have the voting power to marginalize/reduce the industry's latitude for operation (i.e. more regulation under some form of expanded government involvement). Today, those same stocks are rather tame even as work begins on crafting actual legislation.

Again, I think you miss the significance of what was taking place in October 1929 -- it wasn't the exact tariff bill per se, it was the foregone conclusion that the domestic political leadership was ignoring the delicate international situation (with propserity tilted towards the U.S. and reliant upon open markets within the U.S. in order to repay loans) and resorting to old-school isolationism/protectionism. If Smoot-Hawley hadn't passed, the political leadership was obviously in the mood to have crafted something similar. There were already protests and threats of retaliation from other nations as early as the summer of 1929 (which downplays your assertion that the tariff wasn't that big of a deal).

It should also be noted that the hardest hit region and the first major incidence of banking failure that began in Oct-Dec 1930 took place in the midwest in the Federal Reserve District of St. Louis. The midwest/interior is of course where much of the basic industrial and agricultural goods were produced and most reliant upon international trade, so it's of little surprise that by the end of harvest season in 1930 with the tariff in place and unsold goods piling up on both sides of the tariff wall (not to mention a severe drought) that the region's constituents would have difficulty servicing their loans to the banks.
Print the post Back To Top
No. of Recommendations: 2
Any method that uses 10 year data from 1929 would overstate an average P/E because of the huge decline in earnings that begins post WWI and does not trough until 1922. From the trough in 1922, earnings growth is more than fivefold.

What was “normal”? The huge decline in earnings post WWI through 1922, or the fivefold increase the rest of the 20s? At this point, don’t we have PLENTY of historical data that shows aggregate earnings growth follows a path with tremendous, mind-numbing inherent, intrinsic cyclical peaks and troughs (witness the most recent S&P 500 peak earnings versus trailing 12-month earnings) with an overall long-term secular trend growth rate of roughly 6%?

http://www.hussmanfunds.com/wmc/wmc060320.htm

Notice that despite the recent hype, the long-term picture is unchanged. Earnings for the S&P 500 have now moved simply to the 6% long-term growth trend that has connected prior cyclical peaks in earnings for over a half-century (that trend actually goes back much further). Again, everything looks good at the top of the channel.

If this is true (and it is hard to see how it is not from decades of data) then isn’t necessary to arrive at some “normalized” earnings figure in calculating a P/E that captures both boom and bust times? And it seems to me that 10 years is a “good enough” number to use to capture both ends of the business cycle. Anything much shorter and you might be getting a disproportionate share of either a boom or a bust.

I'm also not sure why you declare such a metric to be 'the most reliable measure of market valuation'. From what I can see, such a measure moves rapidly between extremes, does not hover or settle near any one price point and spends varying lengths of time within 'undervalued/overvalued' territory. Especially beginning in the 20th century with the imposition and constant adjustments of the income tax and how such changes impact the value of debt vs. equity, dividends vs. retained earnings, real capital gains etc. Not to mention the near-pervasive inflation that was uncommon in previous measured periods.

Isn’t that the point, right? If a particular valuation metric did NOT oscillate between extremes and did settle near a price point REGARDLESS of subsequent long-term market returns, then of what value is that metric?

I’m a practical guy. In my mind, there is one purpose of a robust valuation metric, and that is to indicate when subsequent long-term (minimum 5 years, preferably 10 years) market returns are likely to be either strong or poor. Take the Fed model which IMO is utterly worthless in this regard. It did NOT correctly identify the tremendous opportunity in stocks in 1982, and it completely missed the danger in stocks at the Oct 2007 peak. In contrast, the Shiller valuation metric both correctly identified that you should have been buying stocks in 1982, and getting the h*** out of stocks in late 2007. I can recall with crystal clear clarity the numerous pundits and strategists making the case that stocks were “reasonably valued” or even “undervalued” in 2007 based on some comparison of forward earnings yield to bond yields. Given the subsequent near 60% peak-to-trough decline (and worst bear market since the 30s) it seems to me like the Shiller valuation model gave investors the right answer on whether to buy, hold, or sell stocks in 2007.
Print the post Back To Top
No. of Recommendations: 2
What was “normal”? The huge decline in earnings post WWI through 1922, or the fivefold increase the rest of the 20s?

Neither was likely normal (whatever normal is defined as)-- but the decline in earnings from the end of WWI is unprecedented, which was the point I was making. The decline in earnings from the end of WWI to the beginning of 1922 was actually larger than the decline from 1929-1932, in both real and nominal terms. According to Shiller's data, real S&P earnings in late 1921 were the lowest on record and nominal earnings were at levels first reached decades earlier.

As such, an average earnings figure incorporating 10 years of data from 1929 includes the years 1919-1921, when earnings are especially depressed. This is turn projects an upward bias into any ratio that uses 10-year average earnings as a denominator. Conversely, there is a downward bias in the 10yr. P/E ratio in the immediate post-WWI period because of the huge spike in wartime earnings.

And it seems to me that 10 years is a “good enough” number to use to capture both ends of the business cycle.

In some 10 year measured periods you might capture as many as 2-3 earnings declines. In fact, Shiller's 10 year method is about to capture such a period as we move into 2010 and the profit peak of 2000 is removed from observation. If S&P earnings of 2009-10 rise but fail to reach the level of 2000, the 10yr. P/E ratio will ironically rise even if earnings are growing and/or the market stays at the current price level.

In contrast, the Shiller valuation metric both correctly identified that you should have been buying stocks in 1982,

... and selling roughly around 1992, give or take a year

and getting the h*** out of stocks in late 2007

I don't see a clear signal in late 2007, in fact the ratio is almost exactly the same in late 2007 as it is in 2003. The whole period from 2003-2007 appears to have the least variation in the entire history of the 10 year P/E.
Print the post Back To Top
No. of Recommendations: 4
Not sure if you noticed this, but your 1930s blog was mentioned on one of the Wall Street Journal's editor's blogs...

http://blogs.wsj.com/economics/2009/06/22/lesson-from-the-de...

Nice to see it starting to get some press. Thanks for all of the hard work you put into it.

Rimpy
Print the post Back To Top
No. of Recommendations: 0
Nice to see it starting to get some press. Thanks for all of the hard work you put into it.

Amen
Print the post Back To Top
No. of Recommendations: 3
This switch by Senator Reed and another Senator is what gave the tariff a majority - and, since he announced it after conferring with Hoover, it's clear that a veto is also not to be expected. So what was the market's response to this? If the mere probability of a tariff bill nine months later was enough for a 50% haircut, surely the inevitability of one in a few days would cause a cataclysm! Dow on Friday ... drumroll please ... up 2.51 points (1.0%).

Just wanted to add one more point on this before departing this subject. I figured something was amiss with your argument so I dusted off my copy of Wanniski's book, The Way The World Works and read what actually happened after Hoover confirmed that he would sign the tariff. If you do not have a copy of the book, you can preview the relative paragraphs (pg.150-151) on Google.

On Sunday, June 15, Hoover announced he would sign the bill and had the major newspapers carry his announcement. There was huge selling of stocks and commodities that Monday morning as the Dow closed nearly 6% lower, its worst decline since .... drumroll please .... October/November 1929, taking the Dow to 230, the exact same level it closed on ..... second drumroll .... Black Tuesday October 29,1929.
Print the post Back To Top