SPX, 1316.63. FV, 1158.43. OFV, 1315.87. POFVF, 1385.86. We note how volatile POFVF is. Therefore, we know it can't be accurate and we are missing something, and this always leads to insight. That said, the sign changes of POFVF from day to day (POFVF(t) - POFVF (t - 1) ) appear very accurate.
The entire strategy the hedge funds are employing is almost trivial: 1) Hedge against Europe disaster. 2) Take advantage of cheap money and borrow from gov at near zero IR and buy equities. This is happening worldwide. 3) Wait for twice a month for mindless morons (Pension Funds, Mutual Funds etc) to come in and buy everything in sight. 4) Hedge Funds sell to the morons in step 3 and go flatish. 5) Go to step 1. This is a house of cards, but it is sold to the masses as "the economy is getting better." In fact, the public is buying stocks expensively from hedge funds because the morons they sell to are only able to act twice a month.
One more thing that I have noticed is that calibration of OFV to SPX is a system onto itself. After the calibration, OFV tracks SPX with great precision throughout the day. I have to look extremely carefully, but 75% of the time that OFV and SPX spread goes greater than two, SPX almost always converges to OFV. I have to look at the risk, otherwise it may be one of those systems that wins 2 handles 75% of the time... So there may be an intraday system in there somewhere. Also, I have been looking to classify which of the dimensions that comprise the FV family best captures the dynamics. I am inclined to try a new idea to further this research, logistic regression on the inputs. I get to use F#, which will make it fun: http://www.elitetrader.com/vb/showthread.php?s=&postid=3429091#post3429091
Seriously now this is all a load of crap. If you truly want to avoid losing boatloads of money, I suggest you consder the problem with being short an index when Apple and IBM report big earnings. That one fact is far more important then all this technical crap you seem mesmorized by. At best for you, markets could do a mild pullback. At worst, this SPX trend will very quickly run up to either the 1350 area or 1370 area. Neither move would be a surprise on a technical point of view. My opinion, without a real negative catalyst, the 1300 level will act as support ( old resistance becomes support ). If 1300 breaks, yes, longs need to bail. But if/until that happens, you don't put real money on it short. Here's why you are losing money. Your personal bias is assigning a much lower value to all markets then they are at. You believe that markets have risen because dumb people are buying. But does that mean anyone who bought AAPL or IBM the last month is dumb ? My opinion at Xmas was that the SPX wanted to test 1350 and TSX back up to 12500, and after that its anybodies guess. Its too late to take that trade, so for me its a wait and see now. The best thing that could occur for me is a pullback. I don't think I'll get it, I've missed the long trade and too bad for me. I was ready to buy the TSX at 11,530 in December and hesitated. Only occured for one daily close and the opportunity was gone. That was also a signal to me there was significant buying support at that level. Last year I sold my TSX mutual fund at slightly under 14,000 based on normal seasonality. I'm a little disappointed I missed an easy 1000 points this fall several times. I'm not chasing, its no longer a great risk/reward. Longer term investors, I think IBM and AAPL are going to shine for 2-3 years. Best entry was prior to earnings. Worst case, they may flatline ( maybe minor downside potential ) in a range if the overall market drops a fair bit. I think that if IBM crosses the psychological level of $200, it could run to $250 and possibly $300 and nobody except current investors will even notice. ps I'm 100% in cash at the moment.
Awesome. Then create CPOFVF for the change in POFVF, which is the prediction for OFV, which is... and if that doesn't work, multiple the signal by -1 to get NCPOFVF.
I think there's been a lot of that multiplying by -1 going on around here by readers of this thread. This is based on the rather untested idea that these numbers mean something, since even to multiply them by -1 they'd have to have some correlation to the course of the markets that is nonrandom. I used to think so, but not anymore. YMMV.
Amazing, this works! I want to see it work to the downside before I get too excited. Hilarious if I end up backing into an intraday system while looking for something that was designed to work on much much higher time frames. I never cease to be surprised...
It is too soon to tell if I have reached some sort of a conclusion, since things are still not completely clear, but I thought I would offer some thoughts that may help others. For those that follow this thread with some hope of getting something useful in their own trading, one thing that has helped me immensely is the journal aspect of it. Clearly, there is nothing in it for me to post here. But I never underestimated the value of having to write down my thoughts, to be able to catalog the trials and tribulations that you go through, ideas that lead to dead ends, etc. It felt like I was trying to put together a precision instrument in a dark room with mittens on. Anyway, FWIW.