Super Short Term S&Ps

eLROY,

in a thread below you said you trade S&Ps in the super short term. What does this mean? I would think to trade super short term you need to be automated. Do you trade the SPY or minis with a time horizon of less than 1 min? what instrument do you use to trade the S&P and what time frame?


Thanks,

Rob

I don't even really use a pit ticker. My theory is that 1) don't get too used to anything which is a byproduct of friction, and 2) the moment somebody starts waving his hands in the pit, long before those tick-punchers on the pier ever have a trade to input, those mini people in the bleachers see it, price it in, and send me a print to tell me about it. Plus, I can't remember the mini's ever having a print pulled.


I can do mini's for $5.00 a round turn, so long as I do them in round lots, even though it is usually split fills. I will be happy to trade in whatever timeframe activity overwhelms the ambient population of trend surfers in that timeframe. So, like, after the Fed moves, the S&P's get very trendy in a sub-1-minute timeframe. Or remember when that plane broke off its tail and went down in Queens, and the news kept dribbling in and dribbling in? That created a series of short-term trendiness explosion-and-decay events.


9/11 created an environment that was so darn trendy in the short term, I'd swear about a million new trend traders got back in business over night. Plus, it revived the trenders in the cash markets, the Island-ECN types. That creates redundancy, and it has been pretty chopped up ever since, thought improving a little bit lately.


Okay, I lost my train of thought, but you get the idea.


Oh, I remember when Timber Hill first started standing in the pit with those high-speed radio-books, placing one-lots. That was about the slowest, crappiest fill I ever got. I could have done better giving my ticket to a hotdog vendor out on Wacker Drive. But today, IB is The King.


eLROY

When you talk about a trend, you're talking about a very very short trend right? not one of those six month moves in a commod. if you've written more about this idea maybe you can point me to the forum and month and i'll look for it.


thanks,

rob

A trend is a correlated body of buyers and sellers. So, suppose it comes across the tape that Tokyo just got nuked. About a million people will try to hit the bid in S&P futures at once. But at once is not really at once. They will all get the news at a different moment.


Or, even if they all got the same news at the same moment, and fired in their orders at the same time using identical equipment, somehow they would end up in a queue in the Globex computer. The tiny period from when the computer matched the first order in its immediate queue to when it executed the last would be a trend. If you were faster than that computer, or had another bid somewhere else to execute against, you could trade that "trend."


But if somebody else were trying to trade that trend at the same time as you, you might hit after them, and hit too deep, after which the market would snap back. In other words, the very shortest trends are actually bodies of like-minded front-runners - or trend traders - smeared out across time. Trend traders themselves, having evolved in response to the same ticker, are correlated.


All phenomena smear out across time. Whether it is analysts contemplating the new numbers from Intel and coming to the same conclusions over a period of time, or more people, earning larger paychecks, and investing them in stocks over a period of time. In other words, in the absence of trend traders, emergent bodies of buyers and sellers in financial markets would be serially correlated.


Let me make this perfectly clear. If it weren't for trend traders, operating on no information other than a series of prints originating from Heaven knows whom or where or why, the Japanese yen would trade at 100 in March, then 101 in April, then 102 in May, then 103 in June. Maybe, by December, some economist would sort out where this buying had been coming from, and why. Heck, weather can trend, diseases can trend.


So trend traders mirror pockets of natural buyers and sellers out there, over different sizes and intra-correlations, and smeared out over different time periods. A lot of S&P day-traders mirror buy-side players who are smeared out across a day they are boxed into. Most currency and interest-rate trend traders are longer term.


The body of trend traders in a given time frame evolves - by the amount of money in their accounts rising and falling - to match the size of the body of buyers and sellers they are fading. When economies become malcoordinated and fall out of wack, and counterparties are missing on one side or the other, these bodies of dislocated buyers and sellers grow in magnitude, and the market becomes "trendy" until the trend traders can make enough money to catch up.


When economies are coordinated, and evenly-matched populations of buyers and sellers arrive at the same rates at the same prices, trend traders chop the prices up and down, without any left-behind buyers or sellers coming in behind them to take handoffs of their inventories. So the price snaps back, as the trend traders self-trigger one another into exiting, and entering in the opposite direction.


I believe the stock markets became less trendy in the intraday timeframe this fall for a number of reasons. For one, I believe more different people started getting involved in stocks again, so volume rose, and there started to be multiple, correlated pockets of buyers and sellers hitting the tape from different diectiosn at the same time - like ripples in a pond from different stones meeting in the same place.


If one trader hits in a vacuum, and trend traders incorrectly extrapolate him to be the tip of the iceberg of a correlated, smeared population, they will whipsaw themselves in a vacuum. Or, if there are multiple, unsynchronized waves hitting at the same time - so that sometimes they harmonize and other times they cancel out - trend traders won't be able to evolve or deduce anything meaningful out of it in any time frame. The hybrid signal will be too complex to disaggregate evolutionarily.


So, for instance, if a market has been trendy in one timeframe, or is expected to be, trend traders will evolve to extrapolate moves in that timeframe. If people fear the market will crash over the period of a day, the waves of people fearing, and then not fearing - and then fearing again - will be smeared out into, say, 10-minute trends. It will be "choppy" in their timeframe - because they are larger than the actual crash sellers they are predicting - but trendy in your timeframe because of them.


Because trend traders chop the market up and down to extremes, natural buyers and sellers will learn to trade counter-trend, in a complementary, or symbiotic relationship. So, you can try to deduce the distance above and below recent prices where buyers and sellers will be waiting. If teh market doesn't go anywhere in either direction all day, maybe they'll all still be waiting, on whatver side they're lurking on, in the last half hour.


Another interesting point to consider is that, of all orders, of all supply and demand, the least visible needs exist among sellers who think the market will go up, and among buyers who think the market is imminent to go down. Both these groups consciously postpone their orders - and inadvertently any visibility of them - into the future...


Enough!


Anyway, so far as other posts, some good ones to start with are listed here:


http://www.twoplustwo.com/cgi-bin/ne....pl?read=35257


Though, really, almost all my posts in the stocks forum have related to trend trading, either directly or indirectly.


eLROY

eLROY,

are you in New York City? Give me an email at papp99@alum.dartmouth.org


Rob