As discussed before, I do not necessarily see price volatility as "risk". I think that is a very sloppy assertion, even for someone with a short term trading horizon. However, without an understanding of the structure of price volatility, and without a definition of significant time horizon, it is difficult to formulate a concept of price risk.A couple of years back I bought and read a book written by Bill Gross. To be honest, I bought the paperback, burned through it on a business trip, and passed it on to a friend so I no longer remember the exact title. But as some of you know Bill Gross is the bond guru at Pimco. Although not particularly sophisticated, the book was a fun read and written in a folksy style much like Peter Lynch's books and I would suggest to anyone interested in the bond market to read it. The theme of the book was that the future was deflationary and we should expect positive, but below average real returns on stocks in the next decade - in the 4-5% range. The book made some interesting points but if you really want to examine the macro-economic slow growth argument you would probably be better off reading Deflation by A. Gary Shilling. Anyway, buried amongst an interesting but arcane discussion about mortgage backed securities, bond futures, and the treasury curve was a two to three page gem called "selling the noise". Those 15 paragraphs made the purchase of the book completely worthwhile to me. It was the proverbial lightbulb thing, ding ding ding - ok now I get it! The concept is that noise is a entity that can be either bought or sold. However, for long term horizons, where we are primarily concerned with the true signal, noise is valueless.Gross' example is the call provision on corporate bonds. In a stable interest rate environment, where interest rates fluctuate +- 2% over multi-year periods, statistics show that call provisions are rarely executed. This is due to the fact that there are a number of costs involved in calling and re-issuing debt that offset the benefits that can be gained from the interest rate spread between the original coupon and the current market rate of interest (presumably lower than the coupon if the call is to be executed). Calling the bonds incurs commissions and service fees; bridge loans to execute the swap tend to have high rates of interest; floating a new bond issue takes time and costs money for underwriting. All of these costs are in current dollars while future interest payments must be discounted. Therefore, two circumstances are required for companies to successfully call and re-issue debt. First, there has to be a substantial spread between current market rates and the existing coupon, and secondly, the issuer must be able to act quickly. These are limiting conditions, especially in a volatile market. Although the market is correct in assigning a risk premium to callable debt, the spread between callable debt and non-callable debt is often much larger than warranted. Therefore, the buyer of the bond is in effect selling a call option to the issuer. Like any option, the more volatility, the higher the premium. However, for long term buyers of the debt, the call provision has nothing to do with the creditworthiness of the issuer, nor is it an issue for interest rate coverage. It is a valueless appendage for the long term buyer interested in the coupon. The coupon is the signal. If you are primarily interested in the signal, why not sell the option. After all, there are worse things in life than a bond being called in a falling interest rate environment!Certainly, higher short term price volatility makes for profitable situations if you can sort out the noise from the signal (longer term trend). Do not underestimate the amount of noise in a financial market. For example, in terms of annual rate of change, if the long term "signal" on an equity is 10% per annum, and the price is changing 3-5% in a week, that is a significant amount of noise. Four percent change weekly is over a 200% annualized rate of change, which is nearly 20 times the baseline signal! From my standpoint,there are three characteristic structures for short term price volatility in financial markets.1. Short term price volatility is anti-persistent with time. It is fundamentally a pink noise process, that is, it is cyclical with no long term trend.2. Cycles of price volatility are non-periodic. There is no average period or amplitude to the cycles of price volatility.3. Cycles of price volatility are asymmetrical. The asymmetry is a reflection on the way information is processed.Antipersistence. Edgar Peters presents an interesting statistical analysis of short term price movements in his excellent book Fractal Market Analysis and demonstrates using R/S analysis (Hurst coefficient) that short term price volatility is intrinsically "pink" noise. Pink noise, as distinguished from white, brown, and black noise, is a term we use in engineering to define cyclical noise that is anti-persistent with time. You can see this anti-persistence in plots of the volatility index - VIX. Non-periodicity. John Bollinger has made a career out of studying price volatility and has even developed a popular statistical method of analyzing volatility with time - the "Bollinger Bands". Bollinger's analysis clearly shows that short term price volatility is cyclical. It rises, then decays in non-periodic cycles. Because the cycles of pink noise are non-periodic, they are generally not amenable to analysis by linear mathematical methods or spectral analysis. They are an artifact of non-linear price mechanisms embedded in the market. Therefore, they are very difficult to time. I think that many market participants have a basic intuitive understanding of the difficulty involved in timing short term price swings or cycles. It is the non-periodic nature of the short term price volatility cycles that makes them so difficult to time.Asymmetry. There are a number of examples of asymmetric pink noise processes in biological systems. The market is inherently a biologically based system. Other than index arbitrage, it is people that decide to buy and sell. And even in the case of programmed arbitrage, no matter how sophisticated the program code, someone still has to set the general parameters for the programs to operate. In general, pink noise cycles in biological systems tend to be compressed on the rising side of the cycle, and drawn out over the decaying side of the cycle. This is because there are reaction and interpretation phases in response to new information.. The reaction phase occurs rapidly while the interpretation phase takes time. The "flight or fight" response is a classic example of the difference between reaction time and interpretation time (or how about the slogan "shoot first - ask questions later" - there is another good example). Human nature 101 creates asymmetry.Cycles create temporal opportunities. However, the non-periodicity of short term price volatility makes precise timing of cycles difficult to impossible to consistently achieve. However, if asymmetry is present in price volatility cycles, then there is a higher probability of making money on the extended side of the volatility cycle. In the financial markets, that would indicate that the more consistently profitable stance is to be contrary when the noise reaches a high level rather than try to time and profit from an increase in the noise level. To be long the noise is difficult because, [a] it is impossible to predict the inflection point from a low noise to a high noise situation, and [b] the time involved on the rising noise line is too short to react profitably to and [c] in the long run the noise is a valueless appendage.Therefore, imho, one should be a seller of high levels of pink noise. This does not necessarily mean a seller of equity and bonds. As shown in the above example, the buyer of callable debt was effectively a seller of noise. This is probably true also of the contrarian equity buyer in many situations. In a panic, sellers will often become price insensitive and lower their ask - in effect buying an immediate put option. The smart buyer is selling that noise. Selling noise is not always profitable, you still are subject to getting clobbered by the disruptive three sigma type events, but it is probably more sensible than buying the noise. It is the approach that the market makers and NYSE members tend to take, and it is the general approach of successful long term investors like Warren Buffett.comments always welcome. good luck.tr
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |
BATS data provided in real-time. NYSE, NASDAQ and NYSEMKT data delayed 15 minutes.
Real-Time prices provided by BATS. Market data provided by Interactive Data.
Company fundamental data provided by Morningstar. Earnings Estimates, Analyst Ra