Stats Vs my Gut feeling

Probability based trading is very new to me and I am seeing lot of interesting quirks. It often contradicts my amateur 'gut feeling' and I am beginning to see how Pros were taking advantage of this crowd behavior.

One example and likely to be in play tomorrow.

We had a big Gap UP today but it's closed decisively by sellers. However we had a higher close but intraday Close < Open. This sort of pattern happens approximately 10% of the times. My usual interpretation is 'sellers took control' of the day. I also assume that if the follow up is gap down the next day, then it's confirmation and will have a bearish bias.

However, stats tell a very different story. We had more than 30 similar occurrences in the past. Win rate for bulls was 65% and profit factor was 2.5. Pretty impressive. It gets even better for gap down open. Totally bizarre.

Please note that this is NOT my primary statistic. I start with Price action like daily HH, HL etc, Gap closures, Seasonality, Vix & Overnight PA. Then I look @ unique factors like above or FOMC, NFP etc.

Bullish stats collide with bearish Tuesday. Curious to find out what happens tomorrow.

June 7 2022 Tuesday Prep

Stats are slightly mixed. Overall bullish except on Big down gap days.

Since we are opening more than 1% which is technically a big gap, with trend and momentum still down, I am hesitant to go all in.

Plan.

I will still enter 50% long and once momentum turns positive, I will scale in 50%

Scale out will be Y-L

June 7 update.

Much better execution today.

Stuck to my long bias based on stats though it looked scary at the open.

50 SPY@ open. Within a minute saw the momentum shifting to long and hence scaled in another 50.

Scaled out 50 @ Y-L.

Brave enough to pick up another 50 @VWAP test and scaled out again @ Y-C.

Though price is going further currently & I am only half loaded, I am glad that I executed my plan well. We are still only 60% of 5 day ATR and ES may go upto 4150. If we pull back to test vWap, then I may scale in again. Otherwise, will exit @ EOD.

I feel like it's slowly coming together for me. i.e. Combining 4 aspects of my trading. i.e
1. Remote past analysis of stats
2. Recent past price action by Market profile & Hi Lo levels.
3. Current battle between buyers and sellers - Delta, Renko & relative volume.
4. Future price movement i.e targets/scale outs -- slow progress.

June 7 review

It's getting better!

Tried to be bit more aggressive in my scale-in aspect of trading.

Yesterday' loss clearly had some recency bias in my first scale out. Y-L was my 1st target but finger itching to take smaller profit sooner that. Having this journal definitely helped as I know I have to explain why I did what I did!.

Once I locked in my 1st profit, then it became a risk free trade. Then the difficult part. I wanted to press higher as my long thesis was working. However, I was too eager and scaled in a bit too early in the Pullback. Though I had to endure a deeper PB, no anxiety there. This time my patience was much better and comfortably waited till Y-C target.

I got even bolder and took my last scale in, as explained in my previous post that I would, at vwap level. Though I was still bullish, my target was conservative as it was bedtime @ different time zone here in Asia. Left my final runner which hit the max target. Perfect! Almost.

The ONE big lesson I learnt from our 'very own' TT was to recognize the game being played sooner and then keep pressing the winner. One has to face the dilemma when having a decent winning day. 'Do you risk turning this into a losing day or much bigger winning day?'. I know what I want now.

Data filters. What do they mean and how to select them?

In medical research, there is a terminology called 'data mining'. This applies when you look for 'something significant' in a clinical study without a clear question in the first place.

Sometimes it's significant. i.e Cardiac side effects for a vaccine (not what you think) in certain age group (20-40) in certain sex group (males).

Most of the times, the finding is spurious and not relevant or useful.

I see a parallel in trading data too. One youtube video yesterday looked at the significance of 'gap down open below yesterday's closing price' in 'June' only. The result was quite bearish. Though it didn't work out, that's not the reason I'm bashing it. I'm curious as to 'why those 2 data points' and 'what's the connection'. May be June is the time when people return from May holiday and behave in certain way. Nope, I'm not sold. If it's a specific day like memorial day or X-mas, then I can see some logic as there may be a behavioral connection. End of the month window dressing makes lot of sense to me with clear behavioural/emotional raionale.

When I use filters in analysing the data, I'm looking for 'predictable crowd behaviour'. e.g. Tomorrow is a futures rollover day. I'm interested in finding out how people reacted on this day. Especially after a bullish close from gap down open.

Overall, slightly bearish and no clear edge. Interestingly, the result was quite bearish after a down day. This correlation persists for another day i.e until the rollover day but not the day after! Does it mean that when the late Bulls are wrongfooted and reluctant to accept the loss are forced to take a bigger loss @ expiration? That makes sense to me (because I have been there)

June 8 2022 Wednesday Prep

Stats are mixed. Slightly bearish but only slightly. No clear edge today.

No clear winner overnight which tends to be bullish i.e continuation move.

Rollover expiry stats are bearish overall but after a bullish close, it is lot less so.

I may rely more on Trend and momentum today. May even stay out for 30 to 60 mins as expecting whipsaw price action.

June 8 update

So far inside day but ONL and ONH quickly taken out.

My discipline muscle was less strong earlier today. Once ONL was taken out and the relative volume was decent, my short bias became strong. I joined the bears & ended up shorting. However, I was more mindful of risk and the lack of statistical advantage. Hence only 1/4th of my usual size. I don't mind being wrong but want to be small when I'm wrong. Plan was to add more if new Lows were made.

It turned out that I was wrong. When the follow up was poor, I realised it's likely locals playing stop hunt. I could have exited @ vwap but I didn't want to be in and out of this market as I expected chop anyways. Hence stayed short and endured the expected ONH to scoop the stops.

I was able to cover my short for a small loss during the deep pullback and then got a better price to short in the 4150 area. Still 1/4th size as my confidence in bears is low but so far they are defending Y-H well. I will bail out on them if that yields. If they make a new Low for the day, then I will scale in.

Good day to observe & learn. I like to observe with a small position

Flat now. Don't like the fact that buyes are testing high of the day again. This could easily rip through Y-H and then fall back to the value or vwap.

Will initiate short if new swing low is made below 4140 and add below 4130.

It's likely to be the end of my day. As expected choppy. Took a couple of shots at the Bulls. Small loss/break evenish day. No complaints. Learnt something.

June 8 2022 Review

Attempted twice to short the market in the AM.

1st one was too soon, FOMO type, and got out for a small loss on a deep pullback.

2nd one was much better but I lost confidence in bears as price hovering near HOD for too long. Risk of ripping higher seemed higher.

3. Final short was executed exactly as per previous entry. Wait for sellers to break the back of Bulls by making a new swing low, LL, and then scaled in after new LOD. Though breakout trades have higher failure rate for me, I felt comfortable as Y-H resistance seemed strong and I considered the probability of new HOD as very low. I did expect price going below 4100 by 10-15 pts with a decent ATR but it was lower @ 55. (5 day ATR 75).

Small losing day turned into a small winning day. Most important lesson was to trade small when I don't have edge & price is chopping. I traded 1/4 size today.

Quirky stats

We closed lower but made a higher low. What's the usual follow up?

I looked at a sample of more than 500 days which suggests the tendency of the market to move lower the following day with Profit factor 0.88.

Then I looked at current market condition, in terms of moving averages, and the price action. This happened on 50+ occasions and the follow up is bit more bearish @ PF 0.7.

Interestingly, had the market made a lower low and lower close, that's a bullish scenario with PF 1.5.

Interpretation - The lower low is the fuel we need to move higher.