Neutral On Monkey Wrench Friday

I’m experiencing a bit of a monkey wrench day plus the tape is a bit wonky – which means that I’ll resort to brevity today. The one chart worth pointing out however is our trusted treasure map:

Just look at the volume profile/distribution and it should be of no surprise that we reversed at the ES 1236 mark (~1239 on the SPX). I told you it’ll be a bumpy ride until we breach 1260 and despite Volar’s skepticism (yes, I know the average has been skewed by 1987 and 2008) I still think we’re looking at a bit o’ whipsaw looming in week 43. Of course IF we manage to push above 1260 the dynamics change completely and after a little shake out I expect quite a bit of follow through.

A word about the ‘trend day alert’ this morning. Obviously I need to fix the alert system (cough cough) but I also wanted to make a point about the data reported today. Scott insists that tick data on the NYSE TICK collected during a 45 minute window is the sweet spot for issuing a trend day flag. Now, I was working on switching things over from a minute to a tick chart last night (yes, ticks on TICK – that’s how we roll) but NinjaTrader was giving me trouble. So I switched it back to a minute chart and then went to bed. What I didn’t realize is that I had left the sample size at 45 minutes instead of 90 – which apparently is a lot more predictive when using minute intervals.

I’ll still have to decide which setting I personally prefer (and maybe some of you guys can help me do some back testing on that) but as I’m a stubborn SOB I of course want it to work both ways. Besides I have heard folks swear on a particular indicator configuration over and over again, just to completely demolish it during basic back testing. Maybe the 90 minute window tests out better – and perhaps a tick window is the way to go. Anyway, obviously the trend day reported this morning is incorrect and just looking at the minute chart it’s clear that had I set it to 90 minutes the alert would have issued a clear whipsaw day (which thus far would have been spot on).

In any case – I’ll test her over the weekend and then give it a trial run Monday and perhaps Tuesday. If everything works according to plan I expect the TrendDay alert system to be up and running by mid week at the latest.

By the way, I had some other ideas that came to me while working on the TrendDay indicator. Compared to that one implementing an InsideDay alert would almost be a trivial effort – maybe for the spoos, the VIX, and the EUR/USD (and other FX pairs). Another good one would be ConsecutiveDays (down or up), FollowThroughDays, VIXOutsideBB, etc. So, the efforts spent on implementing TrendDay will most likely pay off in a big way as you’ll start enjoying a daily series of market meta data.

Hey, maybe I can automate the entire blog, kick my feet up and finally retire while I’m still young and pretty? That would be the day! 😉

Cheers,

Mole