This is a followup to yesterday’s post which can be found here.  In the first post in this series, we discussed the need to have a good set of records and to standardize those records.  Today’s post is aimed at the active daytrader, and takes a look at some of the record keeping issues those of us who do 100 or more trades a day must face.  If this is not you, then hold on for a few days because the third part of this series will actually dig into crunching your numbers in more detail, and that is where it gets interesting.  But first, let’s dig a little bit more into the practical aspects of keeping these records.

First of all, one of the comments on the first post asked how much time I devoted to record keeping each day.  This is a good question that deserves some thought.  The most important thing is that you have a consistent set of records, and that it is within your ability to keep these records religiously.  For instance, I am a fairly active intraday trader, trading sometimes more than 20 names during a single trading day.  On a very active day I will do 50 trades; I rarely do less than 10.  I suppose it would be nice if I had a trade log for every trade that recorded what I was feeling and thinking, what people around me were doing, what the temperature in the room was, how much email I had answered to that point in the day, what related stocks were doing, what the sector was doing, what the market was doing, what related commodities and currencies were doing… you see the list of things that it might be nice to record is basically endless, but I know there is no way I am going to keep those kinds of records.  I may do it for one trade, but whatever I decide to do I have to do for every trade.

So you can see there are two conflicting concerns here — one one hand, you want your records to be as complete as possible.  On the other hand, don’t set unrealistic goals with respect to what you are going to do.  The answer will be different for everyone, but I record the following for every trade when I take it:

• Ticker
• Date and Time (timestamp is automatic)
• Type of trade (approximately 12 categories in my playbook.  this is very intuitive for me.)
• Intended dollar risk per trade

I built a very simple spreadsheet for this that is actually pretty cool.  I type in the ticker and the sheet automatically corrects capitalization on the ticker and then enters a date and timestamp to the left of the ticker in the sheet.  I then have to type in the trade category which is a 4 letter code.  In my case, I basically only trade 3 different setups with a few variations of each.  The letter code for pullbacks is PB.  If it is a pullback to a moving average I enter PBMA.  If it is a high and tight flag, I enter PBHT, etc.  For me, this is very fast and simple, but you must have a system that works with your style of trading.  For me, maybe the most important entry is the intended risk per trade.  It is important that all of these entries are made at the time you enter the trade and you do not revise them later.  For instance, you cannot say you are getting in a breakout trade and risking $100, take a massive loss on the trade, and then revise your entry note to say you were entering a pullback trade and risking $500.  No!  What matters is what you were thinking when you pulled the trigger, not how you revise history to cover your mistakes.  The point of this is to show you what is working and what isn’t, so focusing on your mistakes, no matter how painful, is very important.

At the end of the day, you need a system to consolidate your executions into what I call “plays”.  For instance, imagine this is a simplified chronological trade blotter (price and time would also be needed):

B	300	LVS
B	200	LVS
S	100	LVS
SS	300	AIG
B	300	LVS
SS	100	AIG
S	200	LVS
S	200	LVS
B	400	AIG

I would consider this a transactional level blotter, meaning that it basically shows every time you executed a trade.  However, analyzing this type of a blotter is not very useful at all, because it is obvious that what really happened here was a position was built in both LVS and AIG, and then also both positions were exited in partial pieces.  (In reality, it could be even more complicated.  Perhaps there were two separate long trades in LVS from two different entry signals, for instance?)  A big part of your job at night is to put this together into a format that makes sense.   The way I do it is I have a program that goes through my transactional blotter, first sorts it by ticker and then by time, and keeps a running total of my position for each stock.  Whenever it sees that a transaction would have brought me flat, even for a moment, the system assumes that all of the previous transactions belong together as one “play”.  In my experience, this is something you can only semi-automate because you may have had “nested trades” in the form of additions or perhaps errors where you got flat for a split second that should be combined into the previous play.

You have to decide how to put all of this together for yourself, but for me the end result of this is a single line for each trading “play” for the day that includes the following.  (Note that some of these are pulled directly from my execution blotter, some of them are manually pasted from the records I made when I executed the trades, and some are calculated in the spreadsheet.)

Type of Entry	Manual
Initial Risk on the Play	Manual
Date and Time of first entry	From execution log
Date and Time of last exit	From execution log
Ticker	From execution log
Long/Short	From execution log
# Shares	From execution log
Average price in	From execution log
Average price out	From execution log
Gross P&L ($)	Calculated
Net P&L (est $)	Calculated
Gross P&L as % of Initial Risk	Calculated

Keeping this record is a core trading skill, and something you are responsible for doing every day.  If I kept very good records during the day, I can easily rip all of this info from my blotter, past in the risk and entry information, and be done with the whole exercise in 5 minutes after the close.  If I didn’t keep such good records, then going back through everything and figuring it all out can take an hour or more… which is good motivation to do my job during the day correctly the next day.

I realize as I finish this that this is perhaps the most technical and most boring blog post I have ever written.  I apologize for that, but if you are an active daytrader you should already be using a system like this.  If you are daytrading but aren’t keeping great recrods, then maybe this post gives you some ideas about how to keep those records.  In the next post, we’ll talk about the exciting stuff… how to crunch these numbers to really understand your trading performance.