Fixed link

Inventing Money: The Story of Long-Term Capital Management and the Legends Behind it

Nicholas Dunbar

posted on 10 May 2000

reviewed by Joe McCauley

This is a well-written, informnative and very entertaining book. It begins with doubts that the notion of Adam Smith?s is scientific, presents a brief and interesting biography of Fischer Black, and generally takes up where Liar?s Poker (read it now if you haven?t already) leaves off: with LOR?s idea of portfolio insurance. The information on Merton?s earlier years is also interesting. The CBOT history is described, as well as Black?s love of the CAPM and (one of) the economists? (wrong) notions of equilibrium along with his suspicion of the no arbitrage? argument. Exactly why he was suspicious is left out, but the firming of the no arbitrage idea in the Miller-Modgliani Theorem is hinted at. The history of Ito?s lemma is also stated.

Glass-Steagall and the onset of offshore banking is discussed, as is the creation of hedge funds. The history of LTCM is first-rate. Among the more interesting parts (for me, at least): the connection with UBS and the investment in LTCM by the Bank of Italy, the latter in connection with the ERM crisis in the early ninties. also of interest: how LTCM went beyond sigma? and the lognormal distribution by considering local fat tails (which still ignored extreme large deviations), the emergence of VAR and RAROC as accepted tools in banking and finance, and so-called stress testing?, a theorist?s nightmare but an engineer?s dream. Also stressed is that liquidity was always assumed, which fails during extreme deviations.

Those who have taken the trouble to read Spengler?s last chapters in Untergang... will enjoy Dunbar?s description of the old Arrow-Debreu program, and the philosophically-related program by Merton and his students) to discover/create hidden options, the attempt to capitalize everything everywhere.

There are some points that irritate in a minor way. Feynman?s idea of the path integral is given credit for stimulating LOR?s transition from continouous time Brownian motion to summing over discrete branches on a tree, whereas in reality the path integral applies directly to continuous paths in both quantum mechanics and (as Wiener showed earlier) continous time Brownian motion. In Brownian motion theory drift is confused with diffusion. Statistical independence is confused with randomness. Fama is cited as the originator of the ideology of the efficient market as the underpinning for Bachelier?s theory, and Samuelson is given credit for applying the lognormal distribution in finance. The latter is completely wrong. Osborne (in complete ignorance of Bachlier, it seems) first used the lognormal distribution in finance, whilst Samuelson was still wasting students? time in the classroom lecturing them on the utility and drawing of price as functions of demand and supply that can?t be verified empirically. Newton is mentioned only as a foolish investor in the south Sea Buble, which seems an unnecessary and ridiculous slap. However, options are described as having two sides: fear and greed, and that seems appropriate. Have fun with the book. I did, and I found it to be a strain when more than once I didn?t have enough knowledge of the practicalities of finance to follow certain assertions. Motivation to do more homework ......

PAPER: PostScript only.