DOW started moving well and my bullish bias was executed well for a quick 100 pts. I bought the pullback and expecting fast moves like the last couple of days but YM was in no mood. Then the chop fest began. I am still long 1 lot and tried to scale in twice but bailed out breakeven. My bullish bias persists and considering leaving my position on for a bit longer. I'm NOT adding without seeing a new HOD.


Y

HumbleTrader 

Bread & butter trade in DAX DAX opened gap down, after testing 20 DMA yesterday. Statistically, it's bearish and I had no hesitation in shorting RTH open. Once I received more confirmation, I scaled in once.

Great execution

"Statistically, it's bearish" Could you explain that a bit more? Is the statistic from InvestiQuant? And if so, could you share what conditions you polled the stats for -- something like "gap down", "20 DMA cross", etc?

And is the IQ result a summary panel like what you posted to one of my earlier questions?

HumbleTrader's next chapter Horizontal axis is the timeline. Data starts from 2003. Over the last 20 years, we had around 450 days with these conditions to compare. More detailed info next to the graph now. 1

Great trading, really enjoy the thinking in your journal.

twosigma 

Great execution "Statistically, it's bearish" Could you explain that a bit more? Is the statistic from InvestiQuant? And if so, could you share what conditions you polled the stats for -- something like "gap down", "20 DMA cross", etc?.

Sure. My daily prep starts with 3 main variables.

1. Current market environment - Bullish/Bearish based on 200 and 10 day SMA.
2. Then Yesterday's price action - Higher vs Lower close intraday and relative to previous day.
3. Today's opening price and the size of gap, if any.

Finally, if present, any unique conditions. e.g Opex,NFP, FOMC, Month end etc. and this includes obvious rejection of key levels like MA or swing hi/low.

I have tried many other variables into this basic prep, like VIX, Seasonality, Overnight condition etc. I think they didn't improve the signals significantly and has the downside of reducing the number of data points and hence it's validity.

Now to DAX. It's trading below 10 but above 200 DMA. i.e bullish but weak. Opening price was down and the gap was big. (just an arbitrary number of 50% ATR).

Following graph is the outcome of the theoretical 'long' trades from RTH open to close. You can see the 'average trade' to the left of zero i.e bearish but more importantly, the big left fat tail of >-1% losers.

1


Since gap size of big vs small is subjective, I sometimes have a quick Look at 'any size gap down' and this this is the result I got.

2


It's fairly clear that >1% down days are much common when we gap down (bigger gap, more bearish) and hence I was short and aiming for big gains. However, you can clearly see the risk of +1% up days are also significantly higher. This is where your risk management is handy. Reversal of bias would be even better, should the trend reverse. I was prepared for that after my short but rode the wave nicely and didn't have to open up my toolbox.

Though I thought I milked the trade well when I got out...


d1


..apparently not.

d


I know I can probably never sell the top or buy the bottom but why not keep trying. I just make sure trend is on my side.

I also want to add that the above stats are for US indices. I also use pinescript to get similar and even more granular details. However, my coding and backtesting skills are improving at a much slower pace than my liking. If any pinescripters are interested in collaboration, I would be thrilled.

Until then, I am resigned to the idea that it doesn't matter much. At the end of the day, markets are just a giant mix of human emotions and behavior. Hence, they have more similarities than differences.

Good, Bad and Ugly.

Good - My decision to hold to my MYM till the end of the day turned out to be a good decision. As mentioned in my previous post, I had a good start and grabbed 100 pts but the chop fest went on for too long. EU closing time is 10 PM local time and I generally prefer to be flat by then (pun intended). However, last night I decided to let it stay till the end of US Close due to my bullish price read with supporting data. It worked out well and I almost got out at the high of the day. It's just 1 contract and monetary gain is limited but the intent was to test my thesis and also to get comfortable with 'unmonitored' positions.

Bad - Currently I'm watching DAX going up without me despite having a bullish bias. I put my limit buy order bit lower than what sellers were asking for. Frustrating when you get your thesis right and it's working but without you being part of it. However, the day is still young. Let's see.

Ugly - My Nikkei trade this morning was a short from open. My bearish stats did NOT workout. I had some minor family commitments and should have skipped it. However, I decided to put 1/4 usual size and let it stay till the end of the day which turned out to be a 1% plus day!. However, no harm done as size was small.

P&L is +ve this week, I think.

HumbleTrader 

However, you can clearly see the risk of +1% up days are also significantly higher.

Thanks for the detailed answer.

I've seen the "+1% up" events many times following a gap down with intraday stocks after falling to an ATR-based oversold condition. But it's interesting that it's also strongly present in indexes.

My wish granted by DAX

I was mildly frustrated on missing the opening bullish move by DAX.

My patience and refusal to chase the price paid off. She came all the way back to the opening price to give me a second chance and gladly joined her. I got out for 50+ just under resistance at double top. After a successful retest, I bought the pullback again for another 40+ ride. I'm flat now and probably done for the day until YM opens.

Defining target was tricky today. I recently replied to @1 Minute Man about the challenges in determining them. Today we opened gap up 80 points and hence I had only 120 pts expected range during RTH (total 200 pts recent average). Can DAX have a bigger range? Possible but less likely as the relative volume is low and it's a counter-trend day testing 20 DMA testing from below. However, anything can happen. Main question for me is 'will I regret?'. After grabbing 90 out of 120 points so far, I don't think so.
d

twosigma 

Thanks for the detailed answer. I've seen the "+1% up" events many times following a gap down with intraday stocks after falling to an ATR-based oversold condition. But it's interesting that it's also strongly present in indexes.

You are right. >1% days are not very common event but happened a respectable 25% of days for bulls and 27% for bears in the last 20 years. (Calculated from RTH open to close, without considering gaps). I was surprised that it's higher than my assumption, actually. However, the key is identifying the appropriate market condition for them to manifest.

I lost count of the money disappeared from my account, shorting strong bull markets, expecting 1% down days where the probability is a paltry 6% odds

HumbleTrader 

Following graph is the outcome of the theoretical 'long' trades from RTH open to close. You can see the 'average trade' to the left of zero i.e bearish but more importantly, the big left fat tail of >-1% losers. 1

Thinking more about this... the bimodality of the two tails is striking... and jives with my experience from intraday stock trading that, after a gap down, reversal is almost as common as continuation, as suggested here.

I realize the tail heights are large because they're cumulative (all trades > +1%) versus other bars with a fixed width of 0.1%. Might be interesting to plot the same backtest over a +/- 2% range to see the shape of the bar distribution.

One more thought. This backtest runs from RTH open to close. Would be interesting to backtest the profitability of 2 ATR-based exit conditions.

1) continuation down to -1 ATR re yesterday close
2) reversal and rise to +1 ATR re today's LOD

twosigma 

Thinking more about this... the bimodality of the two tails is striking... and jives with my experience from intraday stock trading that, after a gap down, reversal is almost as common as continuation, as suggested here. I realize the tail heights are large because they're cumulative (all trades > +1%) versus other bars with a fixed width of 0.1%. Might be interesting to plot the same backtest over a +/- 2% range to see the shape of the bar distribution. One more thought. This backtest runs from RTH open to close. Would be interesting to backtest the profitability of 2 ATR-based exit conditions. 1) continuation down to -1 ATR re yesterday close 2) reversal and rise to +1 ATR re today's LOD

Great ideas. Thanks for the suggestions. I am definitely interested to get those granular details. Currently I'm restricted to the presets but will try coding with the SPY dataset available in trandingview.

1. 'Two tail' phenomenon is 'market condition specific' The graph you posted is for weak bullish like DAX. If you consider ES which is in a strong bullish environment, for a 'big gap up open', you get the following. It's much closer to a standard bell shaped curve with a positive skew but still noticeable fat left tail.

1


2. Yes, >1% is cumulative and hence taller but during strong bullish trend, they are still less common than 0.1% increment days. During strong bearish trends, it's extremely common to have big moves. Consider this example below when we have 'big gap down open'. >50% likelihood that we will have 1% days!. I use this data to optimize my position sizing in certain conditions.

2


3. ATR based profit exits and stops is a great way to optimize the system. However, my presets are between 10 to 100% of 5 day ATR and altering the parameters did NOT lead to improvement. It's worth looking at expanding beyond that level. I prefer 150% for trend days and also changing 5 day ATR to 2 days, as you suggested. I will try and connect with you to elaborate these ideas further. Thanks