Related Resources
Essays in Persuasion, by John Maynard Keynes
American Economic History, by Jonathan Hughes and Louis P. Cain
Statistics on the Table, by Steve Stigler
The Speculator
Recent articles:
• We road-test a rule of thumb, 5/15/2003
• Overvalued? Stocks look cheap to us, 5/8/2003
• Market tricks we fall for again and again, 5/1/2003
More...
| |
The Speculator
What's good for bonds is good for stocks
Both stocks and bonds have been doing well of late, a situation that many experts think is at best troublesome and at worst a sign of doom. The numbers tell us the opposite is true.
By Victor Niederhoffer and Laurel Kenner
Heres one for the Dead Myths Department: Stocks cant rise when bonds yields fall, because economic weakness is good for bonds and bad for stocks.
The last several weeks have exploded that story. From March 31 through May 16, stocks rose 11% and bond prices rose 6%.
Since mid-2001, the myth had seemed to be in accord with reality. Bonds rose some 19% as stocks declined 23%. Here are the months in which 30-year bond futures showed a rise of 1% or more, along with the change in stock futures during those months.
| The exception | | Month | Bonds chg | S&P 500 chg | Month | Bonds chg | S&P 500 chg | | July 2001 | 4% | -1% | July 2002 | 3% | -8% | | August 2001 | 2% | -7% | August 2002 | 5% | 1% | | October 2001 | 5% | 2% | September 2002 | 3% | -11% | | January 2002 | 1% | -2% | December 2002 | 3% | -6% | | February 2002 | 1% | -2% | February 2003 | 3% | -2% | | | | Average: | 3% | -3.6% |
|
In fact, as we wrote in a May 17, 2001, column (How the Feds moves shake markets), the relation between bonds and stocks is a subtle one.
Myths are created to explain the unfathomable: things like birth and death. Their function is to keep people content and servile. And the market needs to keep investors servile so that they willingly pay its massive overhead in compensation, equipment and research by trading the wrong way.
Its not surprising, therefore, that the bond-stock myth has given birth to a new, equally awe-inspiring, equally destructive myth guaranteed to put investors on the wrong foot. The new myth is that the United States is headed for a deflationary implosion. With prices constantly decreasing, the story goes, businesses will be unable to make a profit. As in Japan during the last 20 years, decision-makers will be reluctant to spend no matter how much effort the central bank exerts to expand liquidity.
As the story-telling economist Paul Krugman told the tale in a New York Times column at the start of the year: Here's how it can happen: First, for whatever reason, the economy becomes depressed. The central bank responds by cutting interest rates -- but it turns out that even cutting rates all the way to zero isn't enough to restore more or less full employment.
At that point, the economy crosses the black hole's event horizon: the point of no return, beyond which deflation feeds on itself.
Prices fall in the face of excess capacity; businesses and individuals become reluctant to borrow, because falling prices raise the real burden of repayment; with spending sluggish, the economy becomes increasingly depressed, and prices fall all the faster.
We know from Japan's experience that the descent into such a black hole is a gradual process. It seems only yesterday that Krugman and other purveyors of doom were predicting that the coming surge in interest rates would lead to a housing catastrophe that would destroy the recovery and sink the market. A quick Google search shows that we are not putting up straw men; some 56,000 stories mentioning deflation and stocks have appeared in the past three months.
The myth that falling prices will lead to a 1929-style crash has old roots. In The Social Consequences of Changes in the Value of Money, an essay published in a 1932 collection, "Essays in Persuasion", John Maynard Keynes wrote: The fact of falling prices injures entrepreneurs. Consequently, the fear of falling prices causes them to protect themselves by curtailing their operations.
We object to the deflationary spiral story on theoretical, empirical and statistical grounds.
Theoretical grounds
Falling prices are good in the same way that rising prices are bad. Lower prices increase our purchasing power: We can buy more suits and butter with the same income. Businesses can buy raw materials at cheaper prices. Thus, reduced prices represent an increase in the real wealth of consumers and businesses. At the most fundamental level, improvements in wealth are all to the good because they are one of the main components of the pursuit of happiness, held since the Declaration of Independence to be one of the three most important American values.
At a more mundane level, increases in wealth are good because they lead to more spending and more jobs. A good working model is provided by the quantity theory of money:
Given a relatively consistent money supply, price times trade is a constant.
When prices go down, trade (or output) goes up.
Price declines are caused by innovation, improved productivity and vigorous competition -- all highly desirable in a non-command economy.
Empirically
A look at the United States economic history further serves to debunk the deflationary spiral theory. From 1870 to 1900, wholesale prices decreased by about 30% and output tripled.
Economic historians agree that this period showed the greatest relative increase in output and standards of living of any comparably long period in American history. It is no accident that this period of deflation was accompanied and spearheaded by the Industrial Revolution. (This is a good fact to have handy the next time a myth-maker tells you that the robber barons of that period -- the Carnegies, Fricks and Rockefellers -- were rapacious price-gougers. A good discussion of these trends is available in the book "American Economic History" by Jonathan Hughes and Louis P. Cain.)
No matter what we say about the theoretical and empirical facts, the myth-makers are certain to treat them with disdain. Such was the case when we reported on the highly bullish backdrop provided by such things as the low level of supply of stock relative to debt (Why stocks should rise 19% this year, April 20, 2003), the spread between forecast earnings yields and government bond yields, and how no amount of good information can possibly turn any of the bears from their preconceived views.
These columns lead to threats and hateful comments of all stripes directed personally at us; even civil readers reacted negatively. We therefore fully expect readers to write in that our economic analysis cannot be correct, that most economists including numerous Nobel prize-winners are predicting -- dare we say hoping? -- for a Japan-style deflationary depression.
Therefore, we will put our statistics on the table. As Steve Stigler puts it in a passage in the book "Statistics on the Table" we can never quote often enough:
"If a serious question has been raised, whether it be in science or society, then it is not enough merely to assert an answer. Evidence must be provided, and that evidence should be accompanied by an assessment of its own reliability. This test requires more than simple number collection. It involves a careful analysis of the forces that would affect any data, methods for measuring and expressing the uncertainty of the conclusions, and conventions for settling issues such as how much uncertainty is too much, or when an assertion should be rejected and when not."
Our statistics on the table
With our pencils and backs of envelopes at the ready, we begin by noting that it has not always been believed that bond and stock prices move in opposite directions. For 18 of the 20 final years of the old millennium, bonds and stocks moved in the same direction. Indeed, the correlation between yearly returns in bonds and stocks during that period was a very high 75%.
Here is a table showing every occasion in the 1990s that 30-year bond futures showed a monthly return of 2% or more, along with the change in stock futures that month:
| When bonds rise | | Month | Bond chg | S&P 500 chg | Month | Bond chg | S&P 500 chg | | May 1990 | 4.6% | 8.9% | July 1994 | 3.4% | 3.1% | | November 1990 | 4.2% | 5.6% | January 1995 | 2.3% | 2.4% | | August 1991 | 3.4% | 1.8% | February 1995 | 2.5% | 3.5% | | September 1991 | 2.0% | -1.3% | May 1995 | 7.3% | 3.2% | | December 1991 | 5.0% | 11.3% | October 1995 | 2.4% | -0.7% | | May 1992 | 2.7% | 0 | December 1995 | 2.0% | 4.0% | | July 1992 | 4.2% | 3.5% | October 1996 | 3.5% | 2.6% | | January 1993 | 2.4% | 0.1% | November 1996 | 2.8% | 6.8% | | February 1993 | 3.4% | 1.2% | July 1997 | 5.1% | 7.6% | | June 1993 | 3.2% | 0.2% | October 1997 | 2.8% | -3.2% | | August 1993 | 3.6% | 3.4% | August 1998 | 3.7% | -15% | | January 1994 | 2.4% | 3.2% | September 1998 | 3.4% | 7.5% | | | | Average | 3.4% | 2.5% |
|
As the table shows, during the 24 months in the 1990s when bonds registered their big moves, the average move in stocks was 2.5%. During the entire decade of the 1990s, when stocks and bonds were highly correlated, the S&P 500 ($INX) rose 315%.
But that kind of study is descriptive, not predictive. We therefore returned to the 1990s to see what happened to the S&P 500 the month after the 19 months in which stocks rose in connection with bonds (as they have so far in May 2003.)
| Stocks rise, too | | Months when stocks and bonds rose together | S&P 500 return next month | Months when stocks and bonds rose together | S&P 500 return next month | | May 1990 | 1.4% | July 1994 | 3.3% | | November 1990 | 2.9% | January 1995 | 3.5% | | August 1991 | 1.85 | February 1995 | 3.2% | | December 1991 | -2.3% | May 1995 | 2.6% | | July 1992 | -2.3% | December 1995 | 3.2% | | January 1993 | 1.2% | October 1996 | 6.9% | | February 1993 | 2.0% | November 1996 | -1.8% | | June 1993 | -0.7% | July 1997 | 5.7% | | August 1993 | -0.8% | September 1998 | 7.7% | | January 1994 | -3.2% | Average: | 1.1% |
|
Thus, the statistical facts reflect our economic and historical arguments. Far from the new paradigm being the harbinger of doom, it has been a concomitant of boom.
As we write, bonds are up some 5% in May, while stocks are up 2%. It looks like were back to the 1990s kind of model, where bond and stock prices move in the same direction. From a long-term point of view, the 1990s was a very good period for stocks. In the short term, its about 2 to 1 that stocks will go up in June, with the mean change (around which there is much variability) being about 1%.
Final note
We have complete data for bonds and stocks by month from January 1990 available to the present available on our Web site. Our book, Practical Speculation, contains descriptions and explosions of numerous other market myths. Please e-mail us at with other market saws that you suspect might be mythical or would tested. We answer all correspondence.
|
