What the trading styles demand

The same week, four ways 

An instrument opens on Monday and finishes the week somewhere above where it started, having moved up and down along the way. Four accounts hold positions in it. The first opens and closes positions repeatedly inside single hours and by Friday has recorded dozens of results, none of them large. The second opens a position each morning and closes it before the evening, holding nothing while the market is shut. The third opens one position on Tuesday and closes it on Thursday, through two nights. The fourth opened its position a month earlier and still holds it on Friday, having carried it through a scheduled rate decision and a public holiday. The instrument did one thing that week. The four accounts experienced four different weeks.

Those four descriptions are what the word style names, and each is a description of holding period rather than of method. Two traders reading the same chart the same way, applying the same rule to the same instrument, are in different styles if one closes at lunchtime and the other holds into the following month. Everything set out below, the cost, the hours, the exposure to the calendar, follows from how long a position stays open and not from the reasoning that opened it.

The four descriptions 

The four conventional names mark bands of holding period, and each band carries an implied number of positions with it, because the shorter the holding period the more positions a given quantity of activity contains. The two properties are not independent, which is why a style can be described adequately by naming both.

• Scalping. Positions measured in seconds and minutes, many of them inside a single session, each held for a small movement. The count of positions across a week is large and the interval between decisions is short.
• Day trading. Positions opened and closed within the same trading day, with nothing held while the market is shut. A session typically contains a handful rather than dozens, and the account is flat at the close whatever the positions did.
• Swing trading. Positions held from several days to a few weeks, through nights and usually through at least one weekend. Few are open at once and the interval between decisions is measured in days.
• Position trading. Positions held from weeks to months, through scheduled releases, contract rollovers where the instrument has them, and holidays. The number of positions is small and long intervals pass in which no decision is taken at all.

The boundaries between the four are conventions, not definitions, and different sources draw them in different places. A position opened in the morning and closed in the evening is a day trade under every convention. One opened in the evening and closed the following morning is a day trade under some and a swing under others, and nothing about the position itself changes when the label does. What is not in dispute is the ordering. The four names run from the shortest holding period to the longest, and every demand described below runs with them, some rising and some falling.

Cost per unit of movement 

The costs module sets out the charges themselves: the spread crossed on entry, any commission charged per side, and a financing adjustment applied to a position still open past the daily cut off. Styles do not change those charges. What they change is how many times each one is incurred, and how much movement each one has to be measured against. How the gap between bid and ask arises in the first place is set out in the guide on how a spread is formed.

Take the transaction cost first, meaning the spread and commission of one round turn. For a given instrument and a given size that cost is close to constant, whichever style incurs it. The movement a position is held for is not constant at all: it is small by construction in the shorter styles and large by construction in the longer ones. Dividing the one by the other gives the proportion of each movement that the cost consumes, and that proportion is the quantity in which the four styles differ most sharply.

The absolute cost is identical in all three cases. Only its proportion moves, and it moves by a factor of forty between the first case and the third. The two directions matter as much as the three cases: the cost is subtracted from a favourable result and added to an adverse one, so it never offsets and it never reverses sign. That is the break even distance the costs module derives, seen from the other side. A style built on small movements meets that distance on every position, and a style built on large ones meets the same distance far less often relative to the movement it is holding for.

Financing runs the other way across the same four styles. It is incurred per night rather than per position, so it is absent from a style that holds nothing while the market is shut and accumulates on a style that holds for weeks. The shorter styles therefore pay a high proportion of each movement in transaction cost and no financing at all; the longer styles spread one round turn across a large movement and carry a financing line for every night the position survives. Neither arrangement is generally cheaper than the other, because the two cost families are counted against different things: one against movements, the other against nights.

Hours, sessions and the calendar 

A holding period measured in minutes has to be spent somewhere, and it has to be spent while the instrument is worth trading. The sessions module describes where liquidity concentrates and how it thins outside those hours, and a style that depends on opening and closing inside a session inherits that session's clock exactly. The clock belongs to the instrument and to the venue, not to the trader, and it does not move. For an account operated from the Gulf, the hours in which a European or American instrument is most liquid fall in the afternoon and evening, and a style requiring presence through them is requiring presence at those hours specifically, not for an equivalent number of hours at some other time.

The longer styles trade that requirement away and take on a different one. A position held overnight is exposed to every hour in which the market is closed and the world is not, and the earlier lessons on why stops get hit and on guaranteed stops describe what that means mechanically. A stop instruction does not operate while a market is shut. It does not act during the interval in which the news arrived; it acts at the price available when trading resumes, which can be past the level it names. Weekends compound the same point by extending the interval, and the rollover and holiday conditions described in the sessions module extend it further.

The calendar enters in a second way, through scheduled events rather than closed hours. A style whose positions rarely outlive a session meets a scheduled release as an event to be present for or absent from, and either choice is available because the position can be closed beforehand. A style whose positions last weeks meets the same release as something the position sits through, because closing before every scheduled event in a month would leave very little of the month. Neither treatment is presented here as the better one. They are different exposures, and a style commits to one of them by construction rather than by decision.

What a style demands 

Gathering the previous sections together gives a short list of demands. Each is a quantity rather than a preference, each can be stated before any position exists, and each runs in a known direction as the holding period lengthens or shortens.

• Hours of presence, at specific times. Shorter holding periods require attention during the instrument's liquid hours and require it repeatedly; longer ones require far less, but require tolerance of intervals in which a position cannot be acted on at all.
• Decision rate. The number of decisions per hour rises steeply as the holding period falls. Attention is a finite quantity, and a style that consumes it faster than it is replenished degrades in a way the arithmetic of cost does not capture.
• Cost sensitivity. A style built on small movements is sensitive to the transaction cost of every round turn; one built on long holds is sensitive to the financing line and to the number of nights it accrues.
• Exposure through closed markets. Positions held overnight, over weekends and through holidays carry the gap exposure the stop lessons describe. Positions closed at the end of a session do not carry it, and the account holds nothing when the market next opens.
• Aggregate exposure. The portfolio heat described earlier in this module is reached differently by a style holding many short positions in sequence and one holding a few long positions at once, because the second has several positions open simultaneously for weeks at a time.
• Record keeping. The volume of entries a style generates is proportional to the number of positions it opens, which is a real load on the review process the later lessons of this module describe.

Style as a constraint problem 

Style is conventionally discussed as a matter of temperament, as though the question were which of the four is most appealing. The list above is the reason that framing is incomplete. Demands are met or not met, and what meets them mostly sits outside the market entirely: the hours a person is genuinely free, which employment and time zone fix and no amount of preference alters; whether a position can be watched during the hours it would be open; the cost per unit of movement that follows from the instrument and the size traded; and the number of decisions that can be made attentively in a stretch of time.

Read that way, the question changes shape. It is no longer which style appeals, but which styles have their requirements met by conditions that already exist, and the answer is discovered rather than selected. The set of styles that survives that filter may contain more than one, in which case something other than the constraints decides between them, and it may in principle contain none, in which case what the constraints have ruled out is the activity at the size and on the instrument under consideration rather than any particular label.

The practical importance of the distinction is what happens when a constraint is not met. A style whose requirements are unmet is not adopted; it is attempted, and the shortfall shows up as specific behaviour rather than as a general sense of difficulty. Positions opened outside the hours the style assumes, decisions taken while attention has run out, and a holding period quietly reclassified after a position has moved adversely are the recognisable forms, and the last of those is the subject of the final lesson of this module. This page knows nothing about the circumstances of any reader and puts forward no style, no holding period and no rate of activity for anyone.

Where practitioners disagree 

The first argument is whether the four categories carve anything real. One tradition treats them as genuinely different activities that happen to share a vocabulary, on the grounds that the equipment, the attention and the cost structure of the shortest are unrecognisable to the longest. Another holds that holding period is a continuous quantity and the four names are bins with blurred boundaries, so that a trader whose positions last a few hours can be filed under either of two labels without a single thing about the trading changing. The second view is harder to argue with descriptively and the first is what most curricula, this one included, teach from, because bands are teachable and a continuum is not.

The second concerns which demand actually binds. One camp treats cost as the decisive constraint, arguing that once cost is a large proportion of the movement a style holds for, everything else about the style is secondary arithmetic. Another treats attention as decisive, arguing that a cost structure is a term that can be examined and compared while the number of decisions a person can make attentively in an afternoon is not negotiable at all. Both arguments are internally sound, and they are not resolved by evidence anyone has produced, which is why the disagreement is stable rather than moving toward an answer.

The third is whether a style is fixed or adjusted as conditions change. One convention treats it as a fixed container: the holding period is stated in advance and a position that fails to do what was expected of it inside that period is closed rather than reinterpreted, on the reasoning that a movable holding period is not a rule. The other observes that volatility conditions change, that the same instrument can be quiet for a month and disorderly for a fortnight, and that a holding period fixed regardless of conditions is arbitrary in the other direction. The uncomfortable part is that both are describing the same act. A holding period revised because conditions changed and one revised because a position moved adversely look identical from outside, and frequently from inside as well, which is why the fixed convention is often defended on grounds of enforceability rather than on grounds of being right.