Random Comment Regarding Volume Distribution

[Dogpile]

meanth the title to be:

Random Comment Regarding Volume Distribution on ES (got itchy finger and posted too fast)


when you see price building at a single price and suspect the market is balancing.... wait until at LEAST 40,000 contracts trade on ES at a single price before you can expect a potential breakout.

this rule of thumb can keep you fading the moves to the std dev bands until 40k+ contracts trade. somewhere north of 40k contracts, it very often takes 43k+ --- you can stop fading and start looking for a breakout. this simple rule of thumb is just something I have noticed after staring at volume distributions for the last 3 months....

I am calling this: 'Dogpiles Random Rule #1' --- should become a mainstream classic.

seriously though, just watch and tell me if you agree. I am talking about the ES.D volume distribution --- I use Tradestation.

dog

[alleyb]

The only thought I have on volume at a price before a break out occurs is that markets are dynamic and evolving with increasing volume. One day next year I fully expect ES volumes to trade 8 million in one day. Don't get stuck in eternity and also understand how volume changes at moments like option expiry, roll-over, day before/after market holidays plus the traditional so called holiday months like August or December. IE I am saying a degree of flexibility is warranted rather than pure precision.

[Dogpile]

	Quote:
	IE I am saying a degree of flexibility is warranted rather than pure precision.

yes good point.

I meant it to be taken as a concept rather than an everlasting rule. I posted it simply because I wish someone had shared this with me a few years ago. Would have saved me a lot of money not getting chopped up.

The concept is just that you should almost EXPECT the market to play a chop-chop game after some good morning movement (with some exceptions). This choppy action is what is needed to eventually set the market up for an afternoon directional move. 40,000 was just a number I have pointed out as a starting point.

take today as an example (Wednesday):
the market played a sideways chop game for much of the day. 40k contracts did not form at a single price until 3:40pm. Thus, the play was to either fade moves away from 'value' or just wait. Not getting caught up in that chop (or playing fade-trades only before 40k) was a nice guideline to have. The market ended up printing ~55k at 1489.50 today and then had a nice directional move down for -20 pts. Waiting until at LEAST 40k saved you any chance of getting caught in the chop today. Waiting for 55k was golden on this day. Tell me when a market is ready to trend and I can just bracket it and use a stop and clean-up.

Over the last 8 weeks, the median daily PVP is 54,986 contracts. The lowest one that occured is 39,917. As market dynamics change and volume goes up over time in the futures market, 40k will prove too low and I will adjust it upward. But I am just thinking of 40k as a MINIMUM type of number anyway.

[Dogpile]

A few other things I am monitoring:

Extremely high PVP's have become important price pivots.

Multiple daily PVP's building up near each other also have meaning.

Last Wednesday, the daily PVP actually built at a price outside the 'value range' for the day. (ie, the PVP was below the Value Area Low). this PVP was also the lowest (in terms of number of contracts) in the last few months. In this situation, this was a signal of extreme weakness as the market dropped another -30 pts the next day.

I find the 'daily PVP watch' to be something that is quite useful to watch and think about. there is a ton of information there. Would love to discuss what others are seeing and what they may have picked up that I haven't yet discovered, which is surely a ton..

[alleyb]

fair enough re your comment on 40k as a tipping point in volume.
Re: ES
Take yesterday Nov 13th and look at the break down of volume at price and one can see the buying that came in seems to have been roughly 100k net with an average at 1478. Then a quick glance today Nov 14th shows roughly 50k were sold today at an average of 1476 leaving the buyer net long 50k average 1480. So what apparently happened was that the buyers became a little exuberant at todays highs and were forced into chasing their tail. No doubt you will say this is conjecture or subjective and maybe there is an element but if you look at sells vs buys and look at statiscally high levels of liquidity then you can see that yesterday there were several instances of 20k on the way up per price but that today it was a maximum of 14k at one price but most were 10k or less. There is still some selling to be done whether on weakness or on a rally

[alleyb]

as for PVPs there is nothing unusual in that. High volumes have always been the ending and starting of distributions in Stocks and Low volumes in Bonds. since 1982 and 1978 respectively. Prior PVP levels will always act as a magnet and is in fact not the area to trade at as the only advantage is for the broker. PVPs should be viewed as a get out jail card unless of course your objective is to scalp as for some it is for minimum trend returns

[mister ed]

Thats some great observation in there guys. Will have a play with this concept too and see what I can come with.

[Dogpile]

here are some comments on last 4 days. I don't mean for this to be a trading strategy in and of itself -- just a higher level concept to be aware of...

My hypothesis was that the market will generally not sustain a trend until 40,000 contracts trade at a single price. The implication of this is that you can generally think about counter-trend trading until you get to something approaching 40k contracts and then you need to start to be more careful of counter-trend trading as the market might have chopped up enough traders such that a trend could not ensue. The bigger implication is to then gear-up and try to catch the trending move that could develop and just hold on.

Be aware that this is just a tendency I have noticed. The market could trend strongly and THEN do its 40,000 after the movement --- but somehow, some way -- it will get to 40,000 on any decent volume day. I will add the caveat that it is important to figure out ways to confirm this tendency on lower-timeframe such that you do not get 'run-over' in a trend day. Personally, I like to watch how price is reacting to VWAP and to use various standard deviation types of methods to gauge whether the market seems to be acting consistently with my various trading concepts.

Let's look at last 4 days from only a high-level. Tuesday, Wednesday and Friday were very much consistent with my hypothesis - Rangey Price Behavior until 40k.

Tuesday was a trend-day but even that day first printed nearly 50k contracts at 1461.50 before the REAL trend began.

Wednesday was straightforward sideways chop with a giant plunge at the end after printing nearly 55k contracts.

Thursday came close to 40k before dropping really hard (it hit 38k). Obviously, when you see price failing just under VWAP, as it did on Thursday on that test UP, you would want to think short. A market that is rejecting VWAP (while under VWAP) is a short.

Friday was a 'wide and loose' day that saw giant swings but no sustained trend away from VWAP.

All you really have to do is catch 1 of these afternoon moves and your week is made. Or catch just a piece of a few of them and your week is made.

[Dogpile]

Updating this thread for the last few days.

Again, my thesis is that you get a better trending move only after 40k contracts trade at a single price. If price moves strongly before 40k contracts trade, it is likely to ultimately fail and you should consider fading the move on pushes into support/resistance. Once 38-55k contracts trade -- gear-up to play for a potential trending move breakout (bigger win). This is a point of very good reward-risk.

Monday Nov 19th:
We had a 'breakout' set-up coming into this day. Friday had been a narrow (nr7), inside day. The classic Tony Crabel play is to take a first-hour breakout. The market did do a successful first-hour breakout basically as soon as the first-hour was over. When in breakout mode, this is a 'go-with'. That said, note that contract volume did NOT build at a single price sufficiently and thus the move ran out of gas as the odds are that 40k+ contracts WILL build at a price at some point and thus once the market has already trended, you should now EXPECT chop. This is what happened as volume first built at 1442.00 for 48k contracts. This looked like it would be the PVP for the day. The market made a break lower after building that 42.00 PVP and it was reasonable to try for a trending move lower after all this chop. If watching the volume distribution in real-time, you would have seen clearly that buyers came in aggressively at 1437.50 as the volume bar at that price shot out very aggressively indicating extremely strong buying and that price would not go lower. (Watching this in real-time was dramatic and actually also foreshadowed a nice 14-pt move up on the overnight Globex session).

Tuesday Nov 20th:
Tuesday needs commenting on... This day really stands out for being an especially unique, crazy day relative to other days.

1) Volume was heavy on the day with NYSE volume going over 2 billion shares during a holiday week.

2) The volume distribution is quite different than any other day in recent memory. For such a high-volume day, volume was spread out across the entire price range. This is highly unusual. This was the first time in months that no price reached 40k S&P contracts traded while volume for the day was heavy. The range on the day was 34.50 pts.

3) The market showed very strong 'range expansion' off opening price (Opening Price was also 1437.50 -- perhaps not so coincidentally, remember 1437.50 was the huge buying price from the previous day). This upside Range Expansion on decent volume massively failed in an 'outside day'.

Following the rule of thumb for Tuesday would have you doubting the sustainability of a trend until 40k contracts traded (we were not in Tony Crabel type of breakout mode). This was a valid high-level concept because the market tried to trend up and couldn't, then tried to trend down and couldn't sustain it. But this is a bit misleading because we did have a 34.50 pt range on the day --- it was like a day with a trend-day up, a trend-day down, and then another trend-day up all within the same session. This was one of those days that was just very rare and you kind of shake your head at how crazy it was.

Wednesday No 21st:
This was straightforward day where market attempted down before building 'proper' chop. This was a fade. Market then attempted up before building the proper amount of chop and this was a fade. Classic 'Range Trading' until 40k contracts trade. The market then chopped sideways until 45,576 contracts traded at 1431.00. It then faked up and trended lower (sustained its trend). Importantly for intraday traders, you could see how no volume spike occured late in the day -- signalling that big money was not going to hold the market up and that the trend down had legs. This is in stark contrast to Mondays action where on the attempt lower, volume spiked hard at 1437.50, denying the trending move lower.

[snp500]

Dogpile --

That is an excellent indicator on your charts. Is that available somewhere on this forum?

Many thanks in advance!