May 3rd, 2005, 2:02 am
QuoteOriginally posted by: FermionQuoteOriginally posted by: exotiqI do tend to believe that many markets are very efficient (I even hear prop traders complaning to me about them being too efficient and making their jobs difficult), and that contradictory views of technicians will not prevent a reversion to efficiency but instead simply generates noise around what purely rational markets would produce. Note that there is a difference between markets being efficient versus rational; I take the view that many markets are minimally rational, but semi-strong form efficient.If I understand you correctly, you intend that markets are efficient because (a) all the necessary information is available to all players, (b) it is known how to make use of that information, (c) sufficient financing is available to sufficient players to arbitrage any inefficiency and (d) there are no transaction costs, but that some players are irrational and do not make proper use of this information. Is that right?This sounds self contradictory to me (and your distinction between efficient and rational sounds like the difference between six and half-a-dozen) because any noise such irrational players generated would be instantly cancelled by the rational players the irrational players then becoming irrelevant. It also begs the question of which is the chicken and which is the egg. It seems to me that an essential part of the information necessary to trade rationally is prior knowledge of how other players will behave -- which is again either circular (everyone knows how everyone else will trade so there is only one rational way to do so) or incomplete (because some players do not know how other players will behave).My suspicion is that your full set of assumptions is not true (information may not be sufficiently available, but if it were, then it is not sufficiently known how to make use of it) and that the hypothetical efficient market price is not known with confidence except with hindsight even by rational players. However, I suspect that a more rigorous math will uncover better rules for making use of available information (including anticipation of the behavior of other traders) and that those in the know (or the more rational) will be able to profit from it. Even as such rules become widely known and profit opportunities are reduced, transaction costs will necessarily be reduced in parallel and there will still be a vanguard discovering newer rules to scalp ever smaller amounts. The alternative is presumably that markets will no longer be necessary and that some collectively organized body will simply announce the correct prices.No, your summary does not understand what I said correctly. In textbook classification, my views would have markets being minimally rational and semi-strong efficient (and in some ways, you could say hyper-efficient). To me that means:a.) Public information is, just about by definition, available instantly just about anyone that matters, whether or not all of them pay attention to it or understands it.b.) There are players who know how to use the information and push prices instantly towards more or less a "fair value", and other players who overcompensate or misinterpret information and generate noise in doing so. That the fair value are imperfectly known to anyone and that noise generates some degree of doubt and risk of ruin to those who think they are using information correctly means that there is a hazy continuum of players between these two, and who is who is to some degree subjective, but the result being that opportunities to make excess profit per unit of risk using only the publicly available information have been eliminated by competition amongst these players.c.) Sufficient financing is available to enough of the players that matter to leave no arbitrage opportunities behind. The Palm-3Com spinoff has become a textbook example of what should have looked like an true arbitrage, but no longer was once you took into account the financing cost of the stock borrow. One may call this an inefficiency in the borrow market, but if it were a true inefficiency, one would expect to make above-market return per unit of risk by buying the shares (yes, at "high" prices) and lending them out. The markets for institutional financing are overall quite greedy to finance trades when possible, and if financing is not available for what looks like an arbitrage, there is usually a good reason for it.4.) There are transaction costs, which like financing limitations, can create difficult-to-see barriers to what may appear like surface arbitrage. In many ways I might agree that the efficiency of a market has a very strong inverse relationship to transaction costs in that market, and I do consider FX and equity markets to be far more efficient than real estate and life insurance settlement markets. Efficiency is not a black and white result, but rather something that can be compared in an orderly fashion between markets and absolute ideals.For the most part, I try to keep the definition of efficiency a falsifiable one, which I find harder to do in practice than I thought it would be.Unlike in Excel, circular references and chicken-and-egg problems are really not so much a problem in finance, and are less of a paradox than a perspective. You wouldn't say the market for SPX volatility is inefficient or the derivation of its price circular, simply because implied volatility must be backed out as the unknown parameter in our view of future movements, any more than we say it is absurd that YHOO trades at 34.38 because it should be worth 53.47x its trailing 12 month earnings, which we get because the price is 34.38. Markets price at levels we agree to buy and sell at, and the marginal trader on the market is too smart to leave free money on the table.