Decision Moose- Some data

[gap]

Wasn't sure where to put this in existing discussion-


Some data on Decision Moose:
The results shown below are quite remarkable for the period between May 2000 and Today. As mentioned by others the DM philosophy is to pick the best asset class and switch when needed. The asset classes can be mapped on to the HB classes provided you include all the Equity indexes in the Stock class. It takes some courage to invest in just one class and I think the DM author himself does not advise that. As always these results are about the past and one should use care to extrapolate to the future. Regardless, I would strongly advise those interested to subscribe to the DM newsletter. I believe it is one of the best and cheapest newsletters – no BS, no selling and lots of insights and education. (incidentally I have no relationship with DM )
The results shown below are between periods when the selection changes in  DM(not monthly or weekly) although a recommendation is made each week with lots of useful commentary.
Perhaps DM could be used for a portion of VP
                DM PRPFX
Ann. Ret 23% 9%
% of wins 85% 72%
Max DrawDown  -13% -17%
(between switches)
I did not have monthly data so I could not calculate the Std Dev or Sharpe ratio. Others may have it


[img]http://[/img]

[rickb]

There's a detailed analysis of DM at http://www.cxoadvisory.com/2663/economi ... framework/, using the S&P 500 as a benchmark.  The bottom line seems to be that although DM has outpaced the S&P 500 in the past, the advantage it gives seems to be diminishing over time.  It is also not very appropriate for either

a) small portfolios (because of transaction costs)
b) portfolios held in taxable accounts

I would suggest it's not appropriate for a large percentage of anyone's total portfolio either, because of the risk involved in investing in a single asset class.

[stone]

Is it fair to say that it is extremely hard to claim that the DM method has an advantage that is diminishing over time? Everything rests on whether it gets each major event right. Obviously it can never beat stocks when stocks are the favoured asset and DM is in stocks 100%. Basically anything that spotted and avoided the tech crash and 2008 crash is going to look good. What is very hard know is whether it will miss the next plunge. To some extent you could say that there are only two data points to judge it on:  the tech crash and the 2008 crash.

Clive posted about DM a bit too  http://gyroscopicinvesting.com/forum/ht ... ic.php?t=4

Hi Stone.

Decision Moose did OK over the 1999/2000 period http://decisionmoose.com/Moosistory.html

On a total returns basis, a blend of stocks, LTT, gold and cash constantly held for four years is conceptually little different to holding all stocks for one year, all LTT for the next, all gold for the next and all cash for the 4th year.

100% in any one asset however will endure much higher volatility both to the up and down sides.

The trick with RS is to cut losers relatively quickly and run winners. Which is much like a call option (higher risk of a small loss, lower frequency of a large win).

A stop loss strategy might use a fixed stop loss or trailing stop loss. DM in effect uses trailing stops, either by the current leader falling down to below the 2nd best, or (I believe) a 10% trailing stop loss (in case the 2nd best is well behind the leader). DM also uses weekly reviews.

DM in effect tracks stocks, LTT, gold and cash, but with a wider choice of potential candidates for the stock set.

Since 2000 the amount of time DM was exposed to each of stocks, LTT, gold and cash, was something like 30% stock time, 30% bond time, and 20% gold time, 20% cash time. Which if that was held as a constant allocation wouldn't be too dissimilar to a PP. Since 2000 the PP (and that 30/30/20/20 blend) averaged around 7% annualised, whilst the DM annualised 23%. Much of the additional gains that the DM achieved over the PP arises out of periodic large up's from having been 100% in an asset that rose a lot relatively quickly, such as Nov 2001 to June 2002 when it held gold and near doubled in value. Other such big up's, whilst having a relatively tight rein on avoiding big down's, has an overall uplift effect over time. A sort of plateau for a period of time and then jump quickly to a new higher level and then plateau again for a while type motion.

An alternative to DM's choice of relative comparisons might be stocks, LTT, gold, cash and UUP, as generally one of those will tend to always be on the rise and in leading the others would be the asset held, with stop losses set at the next strongest or after a -10% move (in many cases the DM closed out after hitting a stop loss at a higher or comparable price to the purchase price i.e. around a 0% gain or loss). The greatest risk is that of the current holding enduring a hard and fast decline before having the opportunity to stop out. Historically the DM has had a few -20% trades, but countering that its also had more +20% win trades and the odd exceptionally large up win trades, with an overall consequence of 20%+ type overall annualised reward.

As with the Ben Graham approach previously outlined, if rather than solely holding that alone you held perhaps 50% classic PP and 50% Ben Graham PP (or perhaps DM), then the overall rewards might be higher than just 100% pure PP alone. In the case of a 50-50 DM/PP combination perhaps something like 15% total annualised instead of 10% annualised. Which might mean a difference of a 5% real (after inflation) perhaps from the PP compared to a 10% real from the 50-50 PP/DM combination.

Personally I think it is very different to hold assets alternately rather than all at once because holding them alternately means that you forgo any rebalancing gains.

[MachineGhost]

[align=center]

24.86% CAR, -23.02% MaxDD[/align]

Some data on Decision Moose:
The results shown below are quite remarkable for the period between May 2000 and Today. As mentioned by others the DM philosophy is to pick the best asset class and switch when needed. The asset classes can be mapped on to the HB classes provided you include all the Equity indexes in the Stock class. It takes some courage to invest in just one class and I think the DM author himself does not advise that. As always these results are about the past and one should use care to extrapolate to the future. Regardless, I would strongly advise those interested to subscribe to the DM newsletter. I believe it is one of the best and cheapest newsletters – no BS, no selling and lots of insights and education. (incidentally I have no relationship with DM )
The results shown below are between periods when the selection changes in  DM(not monthly or weekly) although a recommendation is made each week with lots of useful commentary.
Perhaps DM could be used for a portion of VP
                 DM PRPFX
Ann. Ret 23% 9%
% of wins 85% 72%
Max DrawDown   -13% -17%
(between switches)
I did not have monthly data so I could not calculate the Std Dev or Sharpe ratio. Others may have it


[img]http://[/img]