Read this book " Rise and fall of Long term Capital management ","when genius failed " by Roger Lowenstein. Had read this once earlier , about a year and half back. Had done a fast read that time. Reread now . Lowenstein deals with the human element more than the technical elements of the story, which is good since he is probably trying to get a larger reader in to the fold. He does deal with some basic explanations on swaps, bond trading, volatility, risk aspects as well as the key aspects of Efficient market hypothesis and random course of securities and equities.
The basic storyline is that John Meriwether, an ex Salomon Bond and arbitrage wing leader , puts together a team of brilliant mathematicians and traders including two future nobel laureates Myron Scholes and Merton Miller to take derivatives tradings to a highly mathematical level through complex models. The major characters involved other than the above are Larry Hilibrand , Haghani and Rosenfield, all math geeks with doctorates from MIT etc to boot.
Their models are built with the following assumptions
1. Markets are efficient and prices respond to information and correctly take in to consideration all knwon factors.
2. Fluctuations ( volatility )in market prices or spreads follow a random course ,and such volatility follows a certain pattern ( bell curve)
3. Deviations from this in pricing , if any is a temporary state and the prices have to return to a state of equilibrium , or in other words follow the past pattern of a ell curve.
The Options pricing model of Scholes, Merton and Fischer in fact rely on a basic premise of Efficient markets and the pricing to follow a certain pattern.
It worked well for a the first three years.The major theme was to locate spreads and mispricing which were not quite in line with their theoretical esimation , book trades in such a way that once the pricing return to normalcy or equilibrium, LTCM can come out with profits.They banked on the fact that inequilibrium ( per their models) can only be temporary and at some point it has to be return to normal. Non adherence for a long time (to their models) was an event which could happen once in say 100 years.
They depended a lot on the convergence in cases of wild deviations and bet on such convergence.
Some of the convergence plays are as
1. Discount differentiel between Bonds based on their interest rate and maturity period .
2.Differentiel pricing between Futures and current price of a security. Make use of differentiel by may be buying futures and selling the current stock ( going short or in other words borrowing ). The expectation is that the Futures prices will go up and hence buying now cheaper makes sense and expectation that the price of securities would get lower
They located mispricings in relation to , one country bond to another country bond with reference to their coupon rates , discount at which they were being sold, one country currency to another country currency etc
Major weakness was that they had tried to model human behaviour .The fund was very higly leveraged. In times of stress or periods which in their terminology , the markets staying in a state of disequilibrium for long periods, there is requirement for additional capital to repay lenders.
To quote Keynes " Markets can stay irrational longer than one can remain solvent"
The had huge investments in Russian bonds , various other slightly risky bonds in which they had invested on the premise that the spread between Treasury bonds and other bonds were far greater than the theoretically right one. Apart from this they had made departures from their own credo of dependence on convergence to certain directional trades in the equity markets. This was very risky and to quote the author, this is like picking nickels in front of a bull dozer.
Russia went back on repayment of debt by devalueing its currency , putting a moratorium on total debt etc. Once markets started seeing red, there was propensity to invest only in safe bonds and the spreads widened. Losses arising this had to be made good by additional infusion of Capital. All the banks get together ( Fed's effort ) and inject capital to bail out LTCM. This was done not only to ensure limiting their losses but also to address the larger issue of protecting investor confidence in the system. The consortium put in $ 3.65 Bio and restricted theor losses. In the final analysis . a dollar put in initially was worth 33 cents. Of course at varying points of time, the invetors got out with profits, huge losses.
The partners who had blind faith in their models were left with very little. Of course as the author says, high finace in the late 90 s, had a knack of rewarding high peformance but not punishing losers. While the partners came donw from the lofty heights , they still retained enough money to come relatively unscathed in terms of retaining a dceent life.
To Scholes and Merton Miller, ,it was a huge blow . It was a case of shaking the very foundation of modern finance theories on options and pricing etc that they had come out with .
These people being the smart cats that they are, still defended and continued to belive that with more time, they would have come out winners.They also partially blamed it a strange set of unusual circumstances which conspired to bring their downfall.
There was also the murky aspect of Goldman , picking up information of LTCM trades while doing due diligence, making use of the information and selling out and making LTCM position get worse.
The Partnership made repeated profits in the first three years, banks in the intial period were servile and obsequious and were too willing to do any trade with them on terms dictated by LTCM. Banks considered it as some kind of honour and prestige to deal with the exalted lot of LTCM. Long term Capital management got away with " no margins on trades", not disclosing the full details of the trades, , keeping each of the two legs of the swap trade with tow different banks to ensure secrecy.
After 3 years of profit, it just took less than a year for the firm to run in to heavy weather and lose the everything.
Funny part is , the losses mentioned are nowhere near the ones sufferred by the players in sub prime crisis of 2008. LTCM collapse almost appears like a honeymoon compared to Sub prime. Same Alan Greenspan was at the helm then and later when the sub prime crisis was gathering momentum.
While there is hell a lot of disclosure requirement and regulations on margins, investment etc on strightfroward loans, and anyway these can very much be gleaned form the balance sheets, derivatives are not disclosed and the risk exposure is much larger. If they had tightened in 1998, they would have possibly had a no or muted sub prime crisis and not the full blown one of 2008
Just to give a perspective, LTCM losses were may be around $ 4 Bio, sub prime was close to $ 1000 Billion.
Do we learn from past ? No, we read up history and the knowledge is history.