For the love of tulips

Many folks appear to be convinced that the current housing situation is akin to any of a
number of other famous financial bubbles of history. Trouble is, those famous bubbles weren’t
very much like what most people seem to assume.

If you talk long enough with someone who’s persuaded that today’s housing market is
experiencing a bubble, the topic eventually is likely to turn to alleged precedents like the
Dutch tulip mania.
Calculated Risk
, the Motley Fool and Seeing the Forest
are among those whose fancy this past week turned to thoughts of those old Dutch tulips.

Much of the popular understanding of the Dutch tulip mania is derived from Charles Mackay’s
delightfully written book, Extraordinary Popular Delusions and the Madness of Crowds ,
whose first edition appeared in 1841. Nine colorfully narrated pages describe the market for
tulips in Holland between 1634 and 1637.

semper.jpg

It was not until 1989, however,
that the record of this episode was really set straight by the careful research of Peter Garber
published in the Journal of Political Economy. Garber documented that the spectacular
tulip prices recounted by Mackay never applied to ordinary tulip bulbs, whose prices, even at
the height of the frenzy, were quoted like produce by the half-pound or pound. The
extraordinary prices characterized instead a few special lines of tulips that had been infected
by a mosaic virus which in some circumstances produced a particular pattern judged to be
beautiful, such as the Semper Augustus shown at the right. But if you liked the effect in a
particular flower, the only way to reproduce it would be to cultivate a few buds from the
original, hoping to have two or three bulbs in the following year with the same qualities.

Unlike the modern notion of a bubble as an asset for which market participants believe the
price will continue to rise, Garber claimed that tulip cultivators always understood that the
price would eventually fall. As long the price falls by less than 50% each year, if the grower
was successfully able to cultivate 2 or 3 new bulbs from the original, he would come out ahead,
because paying 100 guilder for 1 bulb and next year selling 3 bulbs for 50 guilder each is a
winning proposition. In a rational market, the value of the original bulb should equal the
discounted present value of the future net proceeds after cultivation costs of the bulbs that
could be uniquely produced from the original. Given a market for flowers that values unusually
beautiful flowers at a higher price, such a valuation implies that the initial price could be
quite high but quickly fall over time as more bulbs like it become available each year. There
could be (and was) an initial period in which the price actually rises, and quickly, as growers
discover that this particular kind of flower is the one everybody would like to have. But from
there the rational equilibrium is one where everybody understands that the price will fall
rather rapidly, and knowledge of this reality does not discourage a buyer from paying the
initial high price.

And this is precisely the pattern that Garber found, both through the alleged mania and for
the centuries afterward for which more thorough data can be assembled. Garber could find no
indication that anyone experienced financial distress as a consequence of the decline in prices
of some of the fanciest mosaics in 1637, or that the rate of price decline between 1637 and 1642
was significantly different from that seen for other special lines of tulips or hyacinths
developed in the 18th and 19th centuries.

It is true that there evolved in certain Dutch taverns a system for buying and selling the
year-ahead crop for more common bulbs in which more ordinary people participated for more
ordinary sums, in part emulating what they saw going on with the rare bulbs market and hoping to
profit on a more modest scale from the popularity of the flower more generally. The brokers of
such deals were paid with what was described as “wine money,” and such agreements were legally
stipulated to be nonbinding for the 1637 harvest. Garber concluded that, if there was anything
to the idea of a “bubble” in the tulip markets of this time, it applied to these small-time
tavern-negotiated agreements. But Mackay’s account would have been far less entertaining had he
simply reported the details of these rather modest deals.

And if your next question is, “but what about the Mississippi Scheme or the South-Sea Bubble,”
I invite you to take a look at Garber’s examination of these in his paper published in the
Journal of Economic Perspectives in 1990.

I do, by the way, agree with Mackay’s principal thesis that our species is prone to
believing in something that isn’t true just because we see that those around us seem to accept
its veracity. I’m just wondering whether the conviction that there was such a thing as a Dutch
tulip mania might not itself be an example of such a popular delusion.

StumbleUponLinkedInReddit

7 thoughts on “For the love of tulips

  1. CalculatedRisk

    Thanks for the mention! Of course I was just excerpting from The Times article about past manias. I remember reading Garber’s piece: fascinating.
    Luckily we have seen a few bubbles in the last 25 years (most notably the Nasdaq and Nikkei stock bubbles), so we know what they look like and don’t have to rely on any historical inaccuracies. When Garber was pressed in 2000 on the Nasdaq bubble: “I hate to invoke the word bubble,” But he agreed the tech stocks were a bubble: “I thought they were an example. I never saw how competition could be prevented from dissipating the profits that might be there.”
    The key point is bubbles do happen. Housing bubbles look different than stock bubbles, but they do happen and there is significant evidence that one is happening right now.
    Thanks again for the mention.

  2. fatbear

    Fascinating – thanks for the post. For those of us who do not have the benefit of an academic library, is there a link to Gerber’s articles?

  3. J Lonsdale

    I found this argument very interesting, did a bit of checking around and came across this which responds to some of Garber’s points:
    http://www.few.eur.nl/few/people/smant/m-economics/kindleb-rev.shtml
    It seems like partially what is wrong is that he used average prices in his analysis. Garber’s work doesn’t seem to account for the volatility between Jan and Feb 1637. So there may not have been a long bubble, but there was a bubble at least during that time period.

  4. Financial Rounds

    A History Of TulipMania (via Econbrowser)

    James Hamilton at Econbrowser shoots some holes in recent comments comparing housing bubbles to the “tulip mania” of the mid-1600s. It turns out that they really aren’t that similar after all

  5. Dennis Mangan

    The thing I’ve never seen mentioned with regard to the Dutch tulip madness was that it took place during wartime. The Netherlands was at war with Spain from 1621-1648, and the Thirty Years War was going on next door in Germany. That would seem to give an impetus to those trying to find a safe haven for capital: they found it in tulips.

  6. Barkley Rosser

    Garber’s basic argument is that one can never truly know if there is a bubble or not because of the “misspecified fundamental” problem, something the owner of this list has published on also, I believe. However, there is at least one market where this can be overcome with near certainty, that of closed-end funds. There the fundamental is known, the value of the underlying assets, which can differ from the value of the fund. Usually these funds run at a discount, explainable by tax, liquidity, and other reasons. However, when one sees a premium, especially a sharply rising one that then falls, one has almost certainly seen a speculative bubble. Some examples include the closed-end stock funds in the US in 1929 (even if one can play the “misspecified fundamental” game regarding the stock market itself, as Garber does) and many closed-end country funds in 1989-90. Ehsan Ahmed, Roger Koppl, me, and Mark White had a paper on that one in JEBO in 1997.

Comments are closed.