Initial Rebound Probably Near Complete - 2-14-18

I have looked at the market action of the last couple days a few different ways with the idea of what has happened in the past after sizable sell-offs and then the subsequent rebounds.

Today marks 4 days up in a row, and yet price is still below the mid line of the bollinger bands.  So we a have strong rebound after a sharp sell off with wide standard deviation bands.  I have looked at some slight variations of this theme, looking at VIX bollinger bands, and also at multiple up days with the daily MACD still in a down position, and the results are mild to moderately bearish for 1-2 weeks.

One of the scans that was more bearish was 4 days up in a row with today being up 1% or more.  That had a notable bearish skew in forward returns for a couple weeks based on past similar instances.

I also took some time yesterday to compare the most similar past sell offs which then resulted in a failed rally and lower lows for the move.  And the retracement levels were 50-75% of the preceding sell off.  Today hit 50% on the retracement, so we are already in the lower end of the range where past rallies have peaked. 

In follow up to yesterday's post, the VIX/VXV ratio fell back just below 1.0 today.  The past few comparisons of periods of the ratio spiking above 1.0 with a waterfall decline, show that the ratio dropped back to below 1.0 for a few days here and there near the top of short term rallies as prices worked into a retest of the waterfall lows.  So that is still my best idea here, we are near the short term peak of a rebound which is expected to challenge the recent lows.

The VIX also fell 22% today.  I scanned for times when the VIX fell over 20% in a day, and 9 out of 13 had notably greater MAX losses over the next 5 days than MAX gains.  The few instances that did not have greater declines than gains, basically just kept on going up in the context of the recent bull market.  So this suggest we are very near to a point of recognition - either we have had a spike low and price is likely to march higher, or the rally will quickly fail and the next several sessions (1-2 weeks) will have a markedly bearish skew as stocks retest the lows.

As a side note, but somewhat relevant.  I have stated this before on the blog, but price bars with very long tails, like that from last Friday 2-9-18, have a tendency to have retests of the "tail".  I have observed this personally that very wide range bars with long tails are likely to have price come down and tests the lows.

I have heard Steve Nison of candlestick fame, state that on their studies of candlesticks, a successful  "hammer" candlestick bottom reversal has ~60% chance of having price come back to, or below, the mid point of the tail.  So even if a lower low does NOT occur relative to the reversal bar, there may be ~60% chance of price coming back to the mid point of the tail.  That would be ~258 on SPY.

Now the 2-9-18 candlestick is not really a classic hammer candlestick, but in my estimation, a very long tailed reversal with a close towards the upper end of the range, probably has a similar profile.

So my best idea of the odds from this point are for price to come back down to test that 258 region on SPY within the next 2 weeks.  I would not be surprised to see a rally attempt kick in again from those levels.

As another side note, since much of trading is automated computer algos which incorporate key technical indicators and moving averages, I have noticed tendencies for key moving averages like the 200 day moving average to trigger buying programs.  That did occur on 2-9-18.  And if the average is touched again on a retest of the 2-9-18 low, I think it would be reasonable that the buying kicks in again.  BUT, the more times the average is touched without subsequent gap ups and closes above the open, the more that indicates a possible failed rally. 

THEN, if after a couple rally attempts off the average occur, and price is able to break lower and make a lower low, there may be some freefall below where the programs have spent their buying around the 200 day average, and now price rapidly declines to the area of the next significant prior low/high from the uptrend.

In this case, while there are minor lows and highs during the uptrend over the last year+, the last real basing and significant highs/lows were Sept-Nov 2016 in the 208-220 region on SPY.  That seems like a long shot, but if price chops at the 200 day MA and ultimately fails in coming weeks or months, I would not really be surprised to see those 2016 levels be retested on the next move down.

That is looking a long way ahead, but I think it is important to have some view of the risk present at key break points in a market that has been as one sided and speculatively run up as stocks have been in the recent bullish run. 


Pete