The standardized VIX data is flashing a warning signal, possibly a significant warning signal. As of this afternoon (12/4), the S&P 500 is down marginally while the VIX is up more than 3% to just above 17. As you may recall after reading previous posts, it is not the absolute level of the VIX that is important, but it is the standardized level that is the determinant of signals. With this afternoon's increase, the standardized VIX is now above 0.2, which is initial break point I use to determine sell recommendation levels on stocks.
This in of itself is not necessarily the major signal I was talking about. It is the conjunction of this signal and the standardized six-month gains in the S&P 500, which currently stand at just above 1.5 by my calculations. Looking back at the historical data since 2000, there have only been 18 days that showed this dynamic. Just 18 days out of a possible 3,200+ observations. The ominous aspect is that following those 18 observations, the S&P 500 exhibited very poor performance characteristics. The following tables illustrate this point.
The following tables shows the average and median price change following the observations along with the number of observations where the forward performance is positive, the total number of observations, and the percentage of positive performance observations. These observations are for various time frames for both a buy-and-hold case and the when the above mentioned observation is in place.
Buy-and-hold scenarios.
This in of itself is not necessarily the major signal I was talking about. It is the conjunction of this signal and the standardized six-month gains in the S&P 500, which currently stand at just above 1.5 by my calculations. Looking back at the historical data since 2000, there have only been 18 days that showed this dynamic. Just 18 days out of a possible 3,200+ observations. The ominous aspect is that following those 18 observations, the S&P 500 exhibited very poor performance characteristics. The following tables illustrate this point.
The following tables shows the average and median price change following the observations along with the number of observations where the forward performance is positive, the total number of observations, and the percentage of positive performance observations. These observations are for various time frames for both a buy-and-hold case and the when the above mentioned observation is in place.
Buy-and-hold scenarios.
| 1w | 1m | 3m | 6m | 1y | |
| Avg | 0.04% | 0.13% | 0.39% | 0.77% | 1.51% |
| Median | 0.17% | 0.80% | 1.39% | 2.38% | 5.84% |
| # >0 | 1738 | 1865 | 1804 | 1820 | 1909 |
| # Total | 3246 | 3229 | 3188 | 3128 | 2999 |
| % >0 | 53.5% | 57.8% | 56.6% | 58.2% | 63.7% |
Standard VIX above 0.2 and S&P 500 standardized returns above 1.5.
| 1w | 1m | 3m | 6m | 1y | |
| Avg | -2.16% | -3.30% | -0.64% | -0.35% | -12.57% |
| Median | -1.27% | -4.01% | -2.55% | -3.93% | -22.50% |
| # >0 | 5 | 5 | 4 | 6 | 5 |
| # Total | 18 | 18 | 18 | 18 | 18 |
| % >0 | 27.8% | 27.8% | 22.2% | 33.3% | 27.8% |
The significant deviation in the performance should be apparent. However, I think it is prudent to mention one caveat to the observations. Most of the observations fitting the criteria occurred at the market top in or around March 2000. If you remember, this was directly ahead of the Fed draining excess liquidity from the market put in place to prepare for Y2K. (As a side note, I think the liquidity draining events were a main cause of the crash of 2000). This is different world and the Fed is not pulling money from system, but is adding to it. Is this going to make a difference? We shall soon see.
Regardless of the above analysis, the standardized VIX data has tripped a sell point as the model is constructed. In addition, I think the technicals on the SPY look weak, especially noting yesterday's failure in price and volume, a MACD that remains negative, and stochastic that is in a overbought region. I intend to go to a fully short S&P 500 position in the VIX-Trading portfolio.
No comments:
Post a Comment