Why Does Trend Following Not Work for Day Trading

Some people ask why trend following can’t work in the short-term. They want to use trend following on hourly bars, etc.:

I have one more question if you don’t mind: Trend following is generally talked about over the long term, e.g. daily and weekly plus charts. Do you think the lower transaction costs of recent years has made trading a “trend” on a lower timeframe i.e. 60 minute chart possible? The way I see it is that the market is fractal and is “trending” on some time frame or another, e.g. if weekly and daily charts appear to be in a trading range, another timeframe may be “trending”. A few years ago the higher transaction costs meant profiting from the smaller ranges of smaller timeframes unlikely. Now with the advent of lower cost internet brokers e.g. emini you can get 1.60 USD per side, a fraction of the cost from previous years.

Another reader wrote in:

If trend following works on daily charts, why wouldn’t it work on other time frames, such as hourly charts. Assuming one applied the same variables but designed them to work on shorter time frames (units of time become hours as opposed to days), is there any reason why trend following wouldn’t work. I’m specifically thinking of the currency markets which seem to trend on several time frames. Secondly, how does one explain the day trading success of someone like Steve Cohen, who consistently brings in stellar returns day trading certain markets (he charges a 50% participation fee and is very successful).

In theory, all is possible, but we have seen no track record to demonstrate trend following on the short term. You might have some firms with super-large staffs and monster back offices (like Steve Cohen) successfully going for shorter term, but can the average person pull off Cohen’s day trading excellence? Do they have his access and connections? Can the average person overcome the transaction costs of short-term trading?

Ed Seykota View on Short Term Trading

An email question sent to Ed Seykota:

I have a simple question regarding short term trading. You and others have stated that short term trading cannot be as profitable as long term trading since transaction costs eat into profits and magnify losses. It is not because trends do not occur in short time frames as they clearly do. Since markets are fractal in nature, this makes sense. Therefore, if transaction costs could be reduced proportionately such that a reduction in timeframe did not lead to them being a larger part of costs incurred, then short term trading could be just as profitable as long term trading. Is that a correct conclusion to draw?

Seykota’s response:

Check your feelings about wanting to justify your positions by using qualifications and excuses. If pigs had wings, they could fly.

Seykota clarifies later on his site:

Intraday trading is tough since the moves are not as big as for long-term trading and there is no comparable reduction in transaction cost. In general, short-term trading systems succumb to transaction costs and execution friction. You might simulate your system over historical data and notice how sensitive it is to assumptions about where you get your fills. The shorter the term, the smaller the move. So profit potential decreases with trading frequency. Meanwhile, transaction costs stay the same. To compensate for profit roll-off, short-term traders have to be very good guessers. To improve guessing skills, you can practice dealing cards from a standard deck, one at a time. When you become very good at it you might be able to make money with short term trading.

Trend Following Products

Review trend following systems and training:

Michael Covel Trend Following Products
Michael Covel Trend Following Products

More info here.