Dogpile

by the way, there is a hedge fund (Citadel) who apparently has taken the strategy of waiting for inevitible carnage in the hedge fund industry and buying the entire portfolios of blown-up funds at a discount. they are essentially arbitraging the concept that THEY won't be forced to unload mispriced securities since they are well-capitalized... while the hedge fund that is too heavily levered is forced by their loans to sell their mispriced (underpriced) securities (to Citadel).

"The market annihilated the pivot and S/R traders as no levels were respected by the market. Regardless if one is a quant or discretionary trader, the "science" failed." what? many of the hybrid quant/discretionary traders I know had their best month of the year in August. I don't think you can ever say, the 'science' failed. stratgies fail, science does not. for every strategy that fails, there was a strategy that faded the failed strategy. the strategy that worked was based on good assumptions (the science), the strategy that didn't work was based on only a perception of science (bad assumptions). a strategy built on the last X trading years of data might be valid in the future or it might not. traders who had failed strategies simply didn't adjust to the new volatility after an unusually long period of low volatility --- a 'bubble' I would say. if you built your strategy based on the assumption that a bubble wouldn't pop -- and the bubble does pop, I would just call that a bad strategy -- not a failure of science. this is obviously a bigger point but strategies that rely on 'normally distributed returns' and/or 'linear movement' -- and don't properly factor in periods of intense non-linear movement -- will eventually fail. financial markets have been shown over and over again to exhibit 'fat tails'. extreme moves happen more often than one would expect. but this hardly predicts that quant strategies necessarily have to fail -- quant strategies have done extremely well. there is just a point where their own success eventually increases the risk of their own failure as the strategies that are working become widely adopted.

I think we have to distinguish between quant strategies you are reading about in the news and writing quant strategies in Tradestation (or a similar software platform). the quant style to trading (scientific) as a general rule (that is in the news today) is playing off mispriced relationships between securities - not directional movement. I am sure they have 'biases' on directional movement but having read many of the books on LTCM and hedge funds and from talking to those who quantitatively trade options for a business (my brother in-law) -- most of these strategies are simply 'relative value' trades with no directional exposure --- put on in huge size. ie, you buy options that are trading with 28.6 implied vol when the entire options chain on the same security is trading at a 31 implied vol. you instantly hedge it by shorting the underlying. you are letting the market tell you how it is pricing the options (the volatility expectation) and you are buying options at a discount to this and locking in the 'edge'. essentially, you are arbitraging securities whose spreads are wider than the underlying characteristics logically suggest. ie, selling October calls and buying November calls on the same security. I am not saying this strategy is guaranteed to work as it requires 'dynamic hedging' (essentially re-balancing your weightings as changing price can exponentially change the characteristics of a derivatives instrument)-- I am just saying that this is logically sound strategy that is 'expected' to work and exposure to 'shocks' can be neutralized to some degree. LTCM did many monster trades on buying the treasury bond expiring in 29.5 years and funding the purchase by shorting the security expiring in 30 years. hard to logically argue that these securities are much different. let's be clear, good arbitrage strategies work. the problems arise when everyone has the same quant strategy on and everyone is using heavy leverage and are therefore forced to liquidate mispriced securities. this is when things go haywire --- and what happened in 1987, 1998 and last month. now, writing quant strategies in Tradestation or similar programs is generally not about arbitrage -- its about finding small biases regarding location and future direction. this is necessary to understand biases to exploit and make a living off the 'big moneys' crumbs. we have been in the 'information age' for a long time now -- I agree with some recent authors that hypothesize that we are entering the 'conceptual age' --- where computing power is becoming commoditized. in my opinion, trading strategies are optimally executed by augmenting statistical strategies in conjunction with using the right side of the brain too (pattern recognition, synthesizing timeframes, and 'intuition'). but good quant strategies will continue to work -- and 'me-too' quant strategies joining the party late on what 'was' a good quant strategy will clearly have to fail big-time as the best quant strategy-writers will figure out a way to fade their own strategy when its gets too popular. we will be hearing about these guys probably over the next year.

here is one for fun -- this is clearly the exception to the rule -- this is absolutely wild action and should alert you to the fact that something highly unusual is occuring. the market trapped bears, then trapped early bulls. the market then made a final fake down in the final hour but formed a higher low. the market had dropped in the final hour of trading multiple days in a row prior to this. it looked like deja vu -- but the higher low was the key pattern here.

here is another case... you have a mid-day balancing -- and a 'break from balance' --- market is seeking a new level. This is a 'go-with' (the break from balance) -- or at least, don't fight it... Note how the context is different here -- it isn't a price spike at the end of an auction -- it is a price spike less than 2 hours after the initial 'break'.

This is an excellent example of a 'price spike'... yes it could gap down and blow-up the next day -- but it is higher odds that this is a price spike and is just bad location to be short. Getting long takes some guts here and I didn't do it on this day -- but you can 'feel' the emotion involved by investors on this move.

I am continually looking for specific ways to convert the concepts of Market Profile into something tangible. Not necessarily in terms of a set-up – just a way to help understand what the market is doing through something visual. The best way to do this is to use current or recent price action for examples. First the concept: ----- “Perhaps the most important skill you must master to become a successful trader is the ability to distinguish ‘price’ from ‘value’.†Dalton (‘Markets In Profile’ – pg. 100) ----- Last Wednesday September 5, the indices gapped down hard at the open. This must have been a ‘shock’ to many mainstream investors. We were coming off some good, hard sustained buying in previous days. Tuesday was a strongly bullish day (no trading on Monday b/c of Labor Day). Indeed, the gap down was somewhat surprising to me – but it wasn’t a total shock. Dalton continually discusses the relationship between price and value as ‘price’ being volatile and ‘value’ being something that takes time. Sometimes, price just spikes and value never goes there. Price instead returns to value – or instead auctions in the opposite direction, through the last point of value, and in the other direction. Take a look at the first attached chart. Let’s think of VWAP (‘volume weighted price’ for the day) as ‘point value’ and let’s think in terms of # of points above or below VWAP as a ‘value range’. This is different way of looking at ‘value area high’ and ‘value area low’ – but pretty much the same concept. I am using an arbitrary 10 points for the S&P futures. This is not supposed to be a holy grail number – its just a point where it is a lot of points. The market can go 12 or 15 or 20+pts past VWAP – but 10 points is still a lot. More examples to come…

good topic here. not to hijack the thread -- just want to point out a related issue. this seems to happen to me more than it should -- you are trading well and start the day out with multiple winning trades -- you are reading the market well and it just seems so clear. then you enter a position and it does the unexpected and stops you out. now, my first instinct is to either; re-enter because the stop-out was just 'noise' -- or re-enter because you think the stop out represents 'failure' and therefore a position in the opposite direction now makes sense. now you do so and take another position and stop out the second time and have now lost maybe 1/2 or more of your days profit. so its kind of like being on a 'high' -- coming back to earth on the first stop out -- and then trying to get that high back with another trade. This seems to be common and I am not even so sure it is so necessarily bad -- we all face losses at some point -- and after-all, a stop-out does not necessarily mean your next trade MUST be bad. just wanted to share this as this seems to be something I should think about more cause it seems to happen multiple times per week. on the 'failure'/opposite direction trade --- I am trying to make a rule to myself to only enter 1/2 size position because it is not with my original thesis and for this reason, it is likely I just haven't thought it completely through -- the market is moving and you want to catch that days move -- but your mind is BEHIND other market participants -- you haven't deciphered the pattern yet and you are in 'catch back up to the market' mode -- not a good place to be when trading.

hi mark, I exited there for a few reasons. Exits are generally much more complicated than entries but the concepts I was thinking about included: 1. my strategy is generally to look to enter somewhere not too far away from VWAP and find an exit once it has moved a good distance away from VWAP. 2. note that there was a 400 tick buy divergence on the push to 1451.00 3. I was looking for a lower afternoon low --- you can see that this objective was met. net net, the market acheived my objective so I took the gain. waiting for the stop to get hit gives up too much IMO. attached is the buy divergence and a very simple indicator I wrote that marks a dotted line at VWAP +/- 10.00 pts... sometimes it goes less, sometimes more. hope that helps.

here is what I have been using and its worked great... see attached chart for todays action. logic was that if 3 ATR's is a good stop -- then entering on a pullback and using the 3 ATR stop rule will make risk less than 3 ATR's. I have been using 1.5 ATR's as my pullback zone. So in the chart... enter on limit near the lowest yellow line price and put a stop loss in at the lowest red line price -- so your initial risk is roughly 1.5 ATR's (3.0 atr stop with 1.5 atr pullback for entry). This has been working well on a 2-minute chart for me. A good move really shouldn't break the 3 ATR level. fyi, this has nothing to do with a set-up -- this is only the management of a trade once you have decided you like the pattern. man, I am rarely this patient but I really felt like we were going to make a lower low in the afternoon today -- note that it took a full 45 mins before my stop was actually lower than my entry.... but I did get my 'risk' down to 1.5 ES pts after about 20-mins... Easylanguage code for this indicator: value1= c[1]+(avgtruerange(10)*1.5); value2= average(c,2)[1]+(avgtruerange(10)*3); Plot1(Value1, "+1.5 ATRs" ); Plot2(Value2, "+3ATRs" );

Science is having a firm grasp of the statistical tendencies. The art is right-brain pattern recognition -- synthesizing patterns across timeframes in real-time. Very similar to Poker in this regard. There is 1) the math 2) the betting pattern and 3) intuition. Some poker players are good and you don't know how they do it without the math -- but they do. Other players are brilliant mathematicians and get by because they are so good at that - even though they aren't even average at the intution. But most of the top players are very good because they understand the science and the art of the game.

ant, why don't you trade on Fridays?

these overnight plays are lower odds because have to weather a big chunk of the European session which has odds to chop up and down for hours ahead of the Non Farm Payrolls... But reward is large if we trade down as I can then get a free ride on the data release and try for a big win. Pretty high % chance of stopping out -- but big reward potential makes up for this in my estimation. the market should have rallied in the afternoon today if it was internally strong -- and it couldn't. Instead, it made a 'lower high'... this could be a large bear flag forming with continuation tomorrow. that is my hope anyway. if not, then I will stop out and life goes on. as an aside, there is other headline type of risk involving brokerage firms closign their books last Friday and may need to fess up about losses -- or another financial problem circulating (hedge fund losses or bank/mortgage troubles).

went short and plan on holding it overnight for the 'cap' bias.... have worked off the 'pinball buy' and made a 'lower high' in the afternoon... could very well scratch it but have stop in place and hope the cap leads to a gap down. if not, so be it...

might be in a larger bear flag with ABC type of structure... this pattern is bearish and lines up with the 'cap' thesis... to be determined of course but something to be aware of... http://bp1.blogger.com/_5h-SWVGx6Ms/RuAxCR72PLI/AAAAAAAAAaM/uOBfmKr0iOc/s1600-h/ES+abc+up.png

thx ant, I am always too quick on exits to hang on for a bigger win... this is a relatively newer strategy for me... build up a small ES position working the bid on a pullback --- then add with YM as it starts to come out of the flag pattern... I like this entry technique cause you will only have small longs on if it pulls back hard and full positions if it forms a normal type of flag. if it tests up to fill my long and then tanks-- well then I am shit out of luck.... my exits definitely need work.

having a nice start to the month... here are my 3 trades for today. ES/YM did give a pullback but I just scalped it for ~4 pts and 30 ticks respectively. re pinball buy, I thought NQ had best pinball buy set-up so I was long that this morning -- wish I had held it. (note I entered and exited and re-entered -- as you can see on the chart -- held above my original stop so this was a mistake -- as it usually is -- but I hate losing trades to start my day)...

btw, the pinball buy is Rashkes way to mechanize Taylors book. The way the book explains it... you buy a 'test' of the previous days high/low. Today was classic. We had a 'pinball buy' to start the day -- and tested the previous day low. A test can be a higher low or a lower low -- the point is to watch the action 'around' the low for a potential entry. Taylor would call this 'Buy Day, Low Made First' --- its not a long-term trade by Rashkes thinking -- just a trade.

agree, best volume bar of day comes to upside after market in total 'balance' --- could be a trap but odds play is to 'go-with' and manage a stop..

<<How big was the trade target and stop? >> there was no target or stop used. just sell opening price, buy closing price. I ran it also for holding until 'following day close' (2 day hold) but it came out about the same -- it captured more big down days but that was offset by the times it didn't. I am not a mechnical trader. just noticed this one that showed up too -- the Feb meltdown came 2 days after a 'cap'... btw, I have filtered out the days when the trend day down occurs the day after the middle high day -- technically, that is still a 'cap' as 'how far down' the 3rd day goes doesn't matter to technically qualify as a cap --- but I have filtered those days out thinking that the horse has left the barn -- the good kind of cap can be thought of as kind of a slow motion top that has chopped up lots of the over-eager shorts before finally having the plunge... 2 other things: the strategy results are for 2 NQ contracts -- not 1 -- as I always do that due to the small tick value of the NQ e-mini -- otherwise it kind of understates the potential relative to ES.... also, fyi -- those results are for the last 5 years.

added a link now -- Traders Lab just doesn't seem to like that attachment -- I give up after 2 tries -- link is the same thing http://bp3.blogger.com/_5h-SWVGx6Ms/Rt8oxB72PKI/AAAAAAAAAaE/eGMuT36Ob9s/s1600-h/NQ+Cap+Pattern+for+9-6-07+and+9-7-07.png

re-post -- other attachments in post didn't seem to work http://bp3.blogger.com/_5h-SWVGx6Ms/Rt8oxB72PKI/AAAAAAAAAaE/eGMuT36Ob9s/s1600-h/NQ+Cap+Pattern+for+9-6-07+and+9-7-07.png

for anyone familiar with Art Collins book... we have a 'cap formation' that has formed. he swears by this pattern as robust across virtually every futures contract he has tested it on... essentially, its a 3 bar bearish daily pattern. where c<c[1] and c[1]>c[2] and h<h[1] and h[1]>h[2].. see attachment and some strategy results based on shorting the open and covering the close. the key here is just to think of this as a bearish bias over next 1-2 days --- sometimes it needs 2 days to play out -- other times it just doesn't work. also, notice that we have a Taylor bullish bias for tomorrow -- Rashke calls this a 'pinball buy'... in general - I think we have a bearish set-up due to the very solid 'cap' structure and I will be looking for that -- but we might need another day before a bearish bias plays out due to the 'pinball' buy pressure. Just be ready to take advantage of a downside move over the next 2 days should it start to roll that way -- and know that you have potential for a big win with this daily structure.

especially juicy is the fact that both patterns (June 2006 and now) are also 'A-B-C up' patterns on daily timeframe...

agree ant... 'day timeframe trend day up' but profile shape was 'P' -- generally means some long-term selling and is early warning sign that market might be struggling to auction up. though not identical -- might be similar to Daltons example on pages 120-122 (Day 6 in 'Markets In Profile')... despite the late move down, we closed above VWAP and built higher value. there has been no sustained selling below VWAP in 4 days now. as a Taylor disciple -- I will be looking for a 'high to low' day tomorrow -- but trying not to be too biased and just take the set-ups as they come of course... btw, in MIP -- the day after day 6 was a balancing day that built higher value -- then a trend day down --- that would be ok with me...