Return Chasing and Trend Following

[Jack Jones]

http://papers.ssrn.com/sol3/papers.cfm? ... id=2718428

Abstract:     
Return chasing is often cited as one of the primary behavioral foibles of investors, resulting in sub-par returns. Surprisingly, the literature does not provide a generally accepted and testable description of return chasing. This paper proposes a simple definition. It then describes how return chasing so defined differs from trend following and how return chasing explains the shortfall of the returns of active, market timing investors compared to static asset allocation strategies. Finally, it shows that if the trading flows of return chasers are large enough to impact prices, then return chasing provides a powerful explanation of the positive returns earned by trend following strategies, which alternative descriptions of return chasing, such as it is trend following but with too long of a horizon, do not provide.

[dragoncar]

Let me guess: return chasers wait until their current investment is stinking and buy the next investment at a peak.  Then repeat. 

Momentum investors get in at the beginning of the trend and out near the top?

[sophie]

I think of return chasing as sort of like what happens with many drug trials.

Let's say you want to test the impact of a drug on a measure that varies over time (e.g. number of seizures per month or Beck depression inventory score repeated weekly).  You decide to have patients record this measure for 3 months, because you figure that over that period of time the numbers will even out.  At that point you have them try the drug vs. a placebo, and have them continue their diaries for another 3 months.

What ends up happening is that patient selection is biased toward patients who are at their worst during the pre-treatment period.  Inevitably, these patients will "improve" over the next 3 months simply through normal fluctuations.  The "better" times are bound to follow the "worse" times, as the trials are designed to start right after a "worse" time.  This is why so many drug trials show a "placebo effect", i.e. that patients improve both with the drug and with the placebo.  It's not necessarily because of any mysterious inner workings of the mind, just a mathematical oddity resulting from the trial design.

Return chasers get cruddy returns for exactly the same reasons.  A period of outperformance in a particular fund or investing scheme is going to inevitably be followed by a period of poor performance.  During the period of outperformance, everybody thinks that fund manager or investment portfolio designer is a total genius and happily buys in at that time.  You can guess what happens next.

If you really want to chase returns, you want to buy underperforming investments that everybody else thinks you're crazy to buy.  Like, gold, Whole Foods, oil companies.  Of course, you might buy something that really is headed for the dung heap so I have no idea how it would work out.  I'd bet, though, that you'd do a lot better than the investor who plowed everything into stocks in December when the markets looked like a bed of roses, and is now panicking and selling.

[MachineGhost]

What's curious about return chasing is momentum only mean reverts in the 3-5 year time frame, so how the hell can you screw it up by buying what was the biggest winner over the previous year???

[ochotona]

I am going to absolutely hate it    if I get a buy signal* for the SP500 on 3/31/16, what with it being highly overbought. That's just asking for another whipsaw in the next few months. If you buy when overbought, it kind of pointless if the asset quickly takes a dive below the buy price.

I can handicap a buy signal by 1.2%, because on 3/31/15 the SP500 was depressed by 1.2% below trend, but if it goes high enough over the next two days, it will be a positive buy signal.

If that happens, I'm thinking of just skidding the decision forward until the overbought corrects, then taking a reading.

I'm just hoping people take some profits for the next couple of days. Hoping for below 2076.


* one-year total return in excess of T-Bills (0.63% for 1-year Bills)

[MachineGhost]

Short term is always overbought to initiate a bullish signal on higher fractal time frames.

If you try to second guess by waiting for a pullback, you risk missing getting onboard.