Reminds me of the technique that sold when VIX was a certain amount under the 200 day average. That failed in this rally. Stop and think about something: do you really think it's all that important whether the VIX got .08 lower???? I doubt that this technique will stand the test of time....just like most others don't. But I would be interested in knowing whether you backtested this back to 1986? If so, what were the results between 1986 and 1995. OldTrader
TSOKAKIS, Would you mind terribly explaining the basis of your VIX system in words? I am not versed in Amibroker language. The VIX is an powerful indicator and I am always interested in hearing others interpretations of it. Any additional explanation would be greatly appreciated. Thanks.
Schaeffer on Charts: Adding VIX to the Mix By Bernie Schaeffer 8/21/2003 9:25 AM ET Recent action in the CBOE Market Volatility Index (VIX â 19.71) is supportive of the case that the S&P 500 Index (SPX â 1000.30) is about to sustain an upside breakout from its excruciatingly narrow two-and-a-half-month trading range. Note the following from the chart below: The rejection of the brief foray below 20 on July 25 and July 28, which was followed by a surge higher that culminated in the "spike" peak at 25.88 on August 6. The three consecutive daily closes we've now experienced below the 20 level. The "head and shoulders" type pattern over this period, with 20 as the "neckline," that suggests further downside in the VIX. In addition, per the following chart, the 10-day historical volatility of the S&P 100 Index (OEX â 501.89) has just broken below its prior 2003 low of 15.10 percent to reach a new low for the year at 13.29 percent. This tells me that the VIX, which is first and foremost an estimate of the future volatility of the OEX, has room to decline further - perhaps into the mid-teens. Of course, one can certainly make a bearish case for a VIX below 20. And as you can see from the next chart, forays by the VIX below 20 since 1997 have been brief and have occurred at or near market tops. But the VIX environment from 1992-1996 was significantly different, and over that period the low end of the VIX range was 10 rather than 20 (forget the April 1994 low of 5.05, as this was a bad data point). Note also that VIX peaks tended to occur in the 25 zone. My conclusions? The current pattern of VIX price action suggests further downside. Further VIX downside would almost certainly be accompanied by a rally in the SPX. The greater the number of consecutive days the VIX closes below 20, the stronger the arguments in #1 and #2 above. There is no VIX "floor" at 20, as evidenced by the VIX action prior to 1997. On the other hand, brief moves by the VIX below 20 have been consistent with market tops since 1997. The bull vs. bear argument thus reduces to whether post-1997 or pre-1997 VIX patterns will prevail. The heavy focus in the financial media in recent months on the low VIX as a bearish indicator (featured most recently on CNBC yesterday) has bullish contrarian implications. Once a previously arcane indicator becomes widespread in its use, it becomes safe to conclude that the market has already discounted the implications of that indicator. I'd focus on a close by the VIX above 25 as a sign that the 20 level is going to be rejected once again, which would have bearish implications for the SPX. - Bernie Schaeffer
Thanks, Vhehn. I agree... the VIX can get lower and stay low for a long time. I'm looking for an upward trending VIX as a sign of a top, not just a low value. I wouldn't be surprised to see a VIX of 15. I wrote a September SPX put today to put my money where my mouth is.
I use two variations of the 2nd order Infinite Impulse Response Filter[the IIR2], as analytically exposed in Amibroker 4.40 Users Guide, AFL Reference Manual, User-definable Functions. Here is the full code function IIR2( input, f0, f1, f2 ) { result[ 0 ] = input[ 0 ]; result[ 1 ] = input[ 1 ]; for( i = 2; i < BarCount; i++ ) { result[ i ] = f0 * input[ i ] + f1 * result[ i - 1 ] + f2 * result[ i - 2 ]; } return result; } C1=C; LINE1=IIR2( C1, 0.2, 1.4, -0.6 ); k=Optimize("k",0.6,0.1,2,0.1); LINE2=IIR2(C1,0.2-0.01*K,1.4+0.01*K,-0.6); Buy=Cross(LINE2,LINE1); Sell=Cross(LINE1,LINE2);Short=Sell;Cover=Buy; //^VIX,Long +622%,Jan2000 till Aug20, 2003 // Long & Short, +1391% [50%stoploss] The settings for ^VIX are Buy/Sell/Short/Cover at Open, delay +1, commission 0.5%, stoploss -50% The system reached +1800% on May2002, July2002 and Jan2003 ang was a bit below +1000% on May2003 As for the recent performance, from June2, 2003 till now has 7 trades/7winners/0losers with a nice Net Profit +29.5% For the same period the B&H was -11%. Some more details are in the att. gif for Long[below] and Long&Short[above] trades. k=0.6 is the optimal solution. The rest k values give quite good results also and there is no overoptimization here. No need to remind here that ANY past performance of ANY system does not give ANY kind of guarantee for the future performance of the same system.