The Trend is Our Friend: Risk Parity, Momentum and Trend Following in Allocation

[MachineGhost]

The  aim  of this  paper  is to  extend  previous work  by  combining  strategies  in the  context  of  wide ranging multi-asset portfolios. We find that:

•  Risk parity has provided much improved risk-adjusted returns in recent years compared to equal weightings. This appears to be largely the result of the outperformance of bonds. Risk parity adds little to performance within asset classes.

•  Trend following portfolios are a further improvement with higher Sharpe ratios and lower maximum  drawdowns.  This  is  the  case  both  in  multi-asset  portfolios  and  within  asset classes.

•  Adding a momentum filter increases the level of return compared to equal weighting and offers somewhat higher risk-adjusted performance. Momentum portfolios are prone to large drawdowns, however.

•  Combining  momentum  and  trend  following  within  asset  classes  gives  considerably  higher returns than equal weightings, although the risk-adjusted performance is broadly similar to pure trend following portfolios. Adding momentum is thus useful for investors wishing to aim for higher return levels but with an aversion to leverage.

•  A  flexible  asset  allocation  strategy  with  trend  following  that  ranks  all  assets  according  to their  volatility-weighted  momentum  has  shown  a  consistent  high  level  of  return.  This method  has  the  attractive  quality  of  not  requiring  any  asset  allocation  weights  to  be predetermined.

http://papers.ssrn.com/sol3/papers.cfm? ... id=2126478

[stone]

They are basically advocating a "stop loss" strategy such as Clive sometimes talks about. Basically as I see it, it is a way to redistribute your losses through time. Rather than having losses and gains intermingled one after the other, you instead get a series of steady gains until inevitably chance causes you to get repeatedly "stopped out". Lets imagine you "stop out" after a 10% loss. A series of such 10% losses will happen in quick succession sooner or later and that infrequent but catastrophic cumulative loss will erase all of the previous gains. They didn't happen to hit one of those events in their back test. They are basically advocating betting against throwing a six several times in a row at dice irrespective of the odds offered not being good.

If you are dealing with your own money then this strategy is doesn't make sense to me. If you are being paid a performance bonus for gains to someone else's money and you can walk away if you lose it for them and you are a thief - then it is a sensible strategy. That is way it is so widely used .

My impression is that large cap stocks look to whip saw in a totally random way. Perhaps small cap stocks might show some genuine trend. I suppose the weather is something that shows a genuine trend. If it rains today, then it is more likely to rain tomorrow. Does it then always make sense to bet on it raining tomorrow if it has rained today and vice versa? Of course it depends on the potential upside and downside and I think the same thing is true with buying stocks etc. I think the redistribution of risk through time is the overwhelming factor driving trend following.

[Gosso]

http://papers.ssrn.com/sol3/papers.cfm? ... id=2126478

Those are some impressive returns.  The trending strategy makes sense to me.  The major problem I have is that GTAA has not done very well since its inception in late 2010.

http://www.google.ca/finance?chdnp=0&ch ... EARCA:GTAA&

If Mebane Faber cannot get it right then I'm probably screwed.

Also, the returns from Decision Moose haven't been great over the past few years. http://www.decisionmoose.com/

Maybe they're too focused on the short-term and need to look at the bigger picture...

Added: I have been reading through a few of Faber's white papers and I'm impressed.  Maybe it is I that needs to look at the big picture.

[cowboyhat]

If you are seriously considering trend following realize that the profits for your time and emotional energy are equal to the difference in return between a passive style and an active style after accounting for taxes and costs. Imagine how much time you will spend fretting over market trends and your system (my personal experience is lots and lots of time). Ask yourself if you make more money per hour working at your job.

[MediumTex]

If you are seriously considering trend following realize that the profits for your time and emotional energy are equal to the difference in return between a passive style and an active style after accounting for taxes and costs. Imagine how much time you will spend fretting over market trends and your system (my personal experience is lots and lots of time). Ask yourself if you make more money per hour working at your job.

I remember an interesting story I once heard about people who steal parking meters and break them open for the change inside.

The amount of time it took to steal the meter and break it open was calculated and compared to the average take from a typical parking meter theft, and this amount was compared to the hourly wage that the people who empty the parking meters for the city are paid and it turned out that you could make more money legally emptying the meters for the city than you could stealing the meters and illegally emptying them on your own.

[clacy]

That means the parking meters are purely there to generate tickets (revenue) because they are actually a net loser if you just look at the change versus the cost to harvest the change in the meters

[MediumTex]

That means the parking meters are purely there to generate tickets (revenue) because they are actually a net loser if you just look at the change versus the cost to harvest the change in the meters

Not exactly.

Consider that it might take 3 hours to steal a meter and then crack it open and it might net $25.  OTOH, if you worked for the city you might make $10/hour legally emptying dozens of meters throughout the day.  That doesn't, however, mean that the meters you are emptying are collectively making less than $10 per hour in parking fees.

[Xan]

Also, one of the points of the meters is to keep downtown parking turning over.  Even if they only break even money-wise, they still could achieve that function.

[MachineGhost]

I recommend everyon watch this show.  It is really eye opening to see what it is really like at the bottom.  One episode had the mayor of Cincinnati (?) go undercover with the city's parking meter maid who loved her job: http://abc.go.com/shows/secret-millionaire/