Thanks. Is there an easy mechanistic approach on other indicator?
No...fits my data operations so may go produce 10 years of it and chart. Anyone know best charting site/software to create your line/candles and then drop points on top?
BTW I have painstakingly taken Les's awesome daily biotech data and copy pasted by month into giant spreadsheet with text formulas to parse the data back to data. It looks like it may be helpful for bio bottoms and tops. I am mainly using the perecentage above MA data - 100/-100 trends too ubiqitous. Main issue is below 10% on 10D creates a tradeable bottom 90% of time (statistically) BUT the 10% wrong calls kill you (2015/Jan 2016 bear, 2011 flash crash, 2008/09 bear). Being early kills too. Tying less than 10% with other 3 MAs adding to at least 100 reduces samples but increases apparent accuracy to 100% (fewer samples si may be lark). Still has early issue. Trying to tie trade to (1) Below 12%, (2) Bounce to at least X% has issues too as indicator frequently bounces then collapses again in secondary low so you just buy higher, and (3) Initial bio bounces are increduble to behold. Anyway, high 60s due by end of month. Fits seasonals.
Right now based on that work, a low is in process but magnitude of move have me suspecting lower low due next week before seasonals boost into YE. A quant study of three minus 2+% down days w/i 4 days says either crash or big move higher on way (19 samples). Most of crashes came from mkts up over last year - not bios. And we deviated today from 16 samples as we did not get massive gap higher. It says peak on Tue, chop, then up. Target is high 60s by end of month, assuming bulls win. Fits above and seasonals.
BUT the key takeaway for me, is most analogs popping up are flash crash and bear precursors and after-maths. It says definitively we are still in the grip of the bio bear. Lows can be in but trading environment not all clear yet. This is signalled by extreme herding in sector (actually getting more correlated up bounces than long bull runs where A/D is often negative in bios), extreme volatility, lack of trend, and disrespect of price action to key price levels. Ultimately, some key level does hold but just slicing through many others. We bounced today off 50% fib level of November upthrust/triple top from earlier in year/low from August using XBI. This tells me chop may not go lower.
Jon |
