Jason, I absolutely concur that the large specs are profitable in Nat Gas. The small specs are always net long, so they make money in rallies and then give it back. They break even. Given that futures are a zero sum game and the Nat Gas commercial hedgers do hedge forward requirements (ie: lose money), and the small specs are break even, then the large specs must be on the other side of the comm hedgers tradess. How do I know all this? I translate the weekly change in COT positions to $$$$ - sort of market-to-market on a weekly basis. This eliminates/ignores P&L from very short term traders and picks up P&L for position traders. Not perfect but a very good proxy. Cheers
Well we got the pullback to the UVA that I was looking for and got filled. But straight away I knew the trade was in trouble - lots of pressure on the bid. The market tried to bounce but failed at the bottom of the gap. I pulled my second entry order and moved the stop up to 10920. I was happy to eat 15 points because the market was giving clear signs about where it wanted to go and I knew it would be a good day. The action was better on the ES (weaker on a weak day) so I traded that the rest of the day. On the ES I had a neutral bias at the open but T&S, the action, a rising TRIN and a weak TICK(er) was saying LOWER. So took a negative bias (Sell ralies). When we confirmed the gap I sold right at the POC and held on patiently and caught most of the down move. I find this setup to be high probability with an excellent risk/reward ratio. You can use a tight stop and patience is generally well rewarded. I was disappointed that I didn't get a chance to add to my position - I was distracted with IB issues. I figured the market would try to establish a new VA for the rest of the day so that was it for me. MP again held me in good stead today and the action offered up no real surprises. Have a look at the these charts for the YM and ES. IMPORTANT POINT TO NOTE I see alot of people showing charts with lines for the UVA, POC and LVA shown BASED ON TODAYS ACTION. This is clearly misleading because it is postdictive and I'm sure it confuses MP newbies. All my charts show these levels based on the previous days action which is how it should be. I'll check in later, bolter.
bolter, Thank you for the thread. Can a day have more than one POC UVA and LVA? For example two auction ranges, one early in the day and then the rest of the day.... Thanks again.
slacker, Excellent question - you're not living up to your handle. Steidlmeyer and Dalton prescribed calculating a single VA for a days action. However, I find this to be one of the few myopic aspects of the methdology. To me it would seem perfectly natural to have two (or perhaps more) in single session depending upon the action. The classic example is a double distribution where the MP is 'B' shaped. That is, the market establishes value in the morning then breaks significantly, usually leaving single prints in it's wake, and establishes value at a new level in the afternoon. In this case when determining my levels (HVA, POC or LVA) I'll calculate 2 for that day. I'll put the latest VA levels on my chart because they are current. If this is not clear slacker I will post an example on the next double distribution day. There are two other variances from standard practice in calculating VA's that I employ. One I'll discuss at a later date. The other is the standard alogirthm for determing High and Low VA's. Which is, from the POC add the number of TPO's for the two prices above and below, and take the greatest. Repeat the exercise until you have 1 std dev (ie: 68.2%) of the total number of TPO's. I find this gets screwed up when you have action that has single prints in the middle. In this case I will manually adjust theHVA/LVA to best reflect the true value area for the day. The purists among you are probably cringing but I'm a pragmatist. Most importantly I find I get better results. Later
I'm not trading the open today - I have a business dinner. So no trading plan. If I want to trade the afternoon it will have to be Perrier at dinner. Tough call. Anyway here are few things to watch today. Incidentally, these are my levels and will not agree with MP purists. YM POC=10,865 UVA=10,885 LVA=10,845 Nice and symmetrical which implies a bell shaped distribution in this VA. This means the market generally agrees where fair value is. Usually when everybody is agreement then it's time for a change. At this stage it looks to be opening right at the POC (which fits nicely with my theory above). Neutral stance at the open and quickly establish a bias. Watch how it acts around 10895 - strong overhead resistance. Similarly, 10840 is a volume vaccuum which means it may want to back and fill in this area, or it will blow right through. Downside support is way off at 10735. ES POC=1275.75 UVA=1273.50 LVA=1271.00 THE ES is quoted at 1273.25. This is right at a support level and my POC. There is strong overhead resistance at 1278.75, so watch that level. I really don't see any downside support for the ES at this stage. Have fun today guys I'll check in later. bolter
Okay, I do the same thing to track the nat gas winners and losers. It's certainly interesting that the opposite holds true for ES. Good journal!
Thanks Jason. I'll dig out your article. It's hard to find people who really understand the COT data. Cheers
bolter, Those calls were dead on! Does 10840 zone become strong resistance? Time for me to bone up on MP!
Hi bolter, The two most common ways to create a market profile are to: a. use volume * price to create a volume weighted MP b. use 'price alone' to create a distribution of the price data that is not volume weighted. Price distribution (not using volume) is easier to calculate and when using tick data seems to be very usable if not identical UVL, HVL and POC to a volume weighted approach. Do you have any thoughts, you can share, on the 'best way' to construct a Market Profile? Is there another way to calculate MP other than the two approaches described above? After you calculate the levels for the next day are you watching a dynamic MP as you trade? 1 hour MP, 2 hour, total data from market open? Thanks again!
PaperTrader, There is a saying in physics that "nature abhors a vaccuum". This is probably true of markets also - at least that's theory. Sepaking of theories here's mine - Volume peaks form support/resistance, whereas volume valleys/troughs/vaccuums (chose one - I like vaccuum) act like "magnets". If the market has momentum it will blow right through a vaccuum, in fact they tend to accelerate in these areas. However, if the market is sluggish it will tend to get "stuck" around these levels and start filling in the void. To extend the analogy further as the vaccuum starts to fill up the strength of the magnetic field decreases. This is eactly what's happening on the YM this morning. A fundamental premise of MP is that trading/activity/volume will generally be distributed "normally" (ie: bell shaped) around what is generally considered to be "fair value" by market participants. This is so on an intraday basis as it is over weeks and months. The simple force at work here is that traders will sell markets that are above fair value and buy those below fair value. Enough of theory - it gets a bit spooky. Have a look at my latest CUMULATIVE volume chart (since 1/1/06) for the YM. I'll be damned if this isn't the perfect H&S pattern!!! Long term value for the YM is around 10895. Unless there is a dramatic shift in the long term perceived fair value for this market then it should (SHOULD!) fill this profile out until it resembles a normal distribution curve. Of course it nevers happens exactly like that, but the key thing to understand is that that's what an auction market is always attempting to achieve. Hope this fills in some blanks - if you'll pardon the pun. bolter