Seeking Alpha

varan's  Instablog

varan
Send Message
Individual investor.
  • An Adaptive Asset Allocation Example 17 comments
    Mar 20, 2013 3:13 PM

    This is a quarterly rebalancing strategy that works well to minimize portfolio volatility without any significant deleterious effect on the returns.

    1. The basket consists of two parts (a) a set of assets nominally correlated with the market, typically equity ETFs and (b) a set of assets which have low or negative correlation with the market, typically fixed income ETFs. The number of fixed income assets should be as many as or preferably more than the number of equity assets.
    2. The rebalancing schedule consists of three tranches:

      (a)at the beginning of every quarter choose a set of assets that performed the best during the immediately preceding month,

      (b)at the beginning of every half-year choose a set of assets that performed the best during the immediately preceding month, and

      (c)at the beginning of every year choose a set of assets that performed the best during the immediately preceding quarter.

      Obviously the assets from the three tranches can be commingled at the beginning of each quarter to be equally divided in the three trenches. Further the number of selected assets should be close to the number of equity ETFs in the basket (as in 1(a)).

    As an example we consider a modified version of The All Weather Portfolio (although any combination of equity and fixed income assets works well) described here: http://seekingalpha.com/article/1289651-allocating-within-the-all-weather-sailboat-portfolio-part-2 . My calculations (many of which span periods starting in 1991) suggest that any basket satisfying the criteria described above will yield similar or better results.

    The basket consists of the following:

    IVE IJS IJJ EFA EEM TIP

    IPE TDTF SCHP PRF VCIT LQD

    AGZ MBB TLT PLW VGLT SCHR

    At each update for each trench, six assets were chosen.

    Here are the results for 2003:2013

    CAGR 12.85%

    Sharpe Ratio .86

    Kelly Fraction .41

    Monthly One Factor Alpha .54%

    Monthly One Factor Beta .54

    Monthly One Factor R^2 .56

    The equity curve, the Manhattan asset allocation diagram and the basic asset allocation diagram are shown below.

    (click to enlarge)(click to enlarge)(click to enlarge)

    Disclosure: I am long IJS, LQD.

Back To varan's Instablog HomePage »

Instablogs are blogs which are instantly set up and networked within the Seeking Alpha community. Instablog posts are not selected, edited or screened by Seeking Alpha editors, in contrast to contributors' articles.

Comments (17)
Track new comments
  • Learner16
    , contributor
    Comments (105) | Send Message
     
    I guess it helps we had a bull market in bonds during that time. Would it be possible to see the equity curve including a bear market in bonds?
    20 Mar 2013, 08:20 PM Reply Like
  • varan
    , contributor
    Comments (3521) | Send Message
     
    Author’s reply » The premise is that if there is a bear market in bonds, there will be a bull market in stocks. For the last three quarters, the amount in bonds has been quite low, in keeping with the recent trend. Perhaps we can add a few assets that appreciate when both bonds and stocks are in a free fall.

     

    There are two basic ideas: (a) have enough assets that are uncorrelated, and (b) average across multiple holding periods and ranking periods. The latter is not often emphasized.
    20 Mar 2013, 08:27 PM Reply Like
  • Learner16
    , contributor
    Comments (105) | Send Message
     
    Thanks, Varan, for the reply. The trick is that if bonds and equity are long-term uncorrelated, there will be some periods of time where both will be down. Maybe we could add cash as a third uncorrelated asset?

     

    I agree wholeheartedly with your point b). Averaging across multiple holding periods and ranking periods seems to me as important as using non-correlated assets.
    21 Mar 2013, 10:46 AM Reply Like
  • varan
    , contributor
    Comments (3521) | Send Message
     
    Author’s reply » Perhaps we can add IYR, GLD, and SHY. This will encompass all sorts of conditions.

     

    The main purpose was to show an example of adaptive allocation. As you can see from the first allocation diagram, the strategy works as you would want it to - changing the allocation to equity ETFs according to market conditions, and responding sufficiently rapidly to the dynamics of the market to avoid large drawdowns.
    21 Mar 2013, 12:27 PM Reply Like
  • MrBobDobalina
    , contributor
    Comments (70) | Send Message
     
    Hi Varan - I need a little clarification here.

     

    There's just one basket of etf's to choose from, right?

     

    From that basket, you have three time periods -

     

    beginning of each year, you choose the top 2 performers for the previous quarter.

     

    Jan1 and July1 (each half year) you choose the top two etf's that performed the best during the previous month.

     

    And the beginning of each quarter you choose the top two etf's that performed the best during the previous month?

     

    Is this correct?

     

    From the graph I assume your're showing the 14.3% and 14$ as the two largest drawdowns? Those are based on month end results, not daily, right?

     

    Thanks!
    21 Mar 2013, 11:57 AM Reply Like
  • varan
    , contributor
    Comments (3521) | Send Message
     
    Author’s reply » Conceptually, think of this as three separate streams which allocate weights to all the ETFs. For each stream the weights are computed by selecting the top six (not two) during the evaluation period - so allocation will be 1/6 for the selected ETFs, and zeros for the others.

     

    For one stream the weights change every quarter. For the other the weights change every six months. For the third the weights change every year.

     

    So at any given time you have three sets of weights. On the first trading day of each quarter average the three sets of the weights for allocation for the quarter.

     

    The drawdowns are based on monthly data, and only the drawdowns larger than 10% are shown in the figure. In this case there are two drawdowns of 14%. The next highest drawdown (not shown) would be less than 10%.
    21 Mar 2013, 12:20 PM Reply Like
  • Matts_options
    , contributor
    Comments (4) | Send Message
     
    Varan, thanks a lot for this blog and particularly this entry, with the idea of average across multiple holding periods and ranking periods. I am studying some of your methods along with others on the web and comparing the results I get... Keep up the good work!
    25 Mar 2013, 01:29 AM Reply Like
  • Alpha Man
    , contributor
    Comments (208) | Send Message
     
    Varan, I really enjoy your articles, good stuff!
    3 Apr 2013, 03:05 PM Reply Like
  • varan
    , contributor
    Comments (3521) | Send Message
     
    Author’s reply » @Alpha and @Matt

     

    Thanks a lot for your interest and kind words.

     

    I write these mainly to record the results that I find interesting. Positive feedback from the readers is an added bonus that I greatly appreciate.
    3 Apr 2013, 04:35 PM Reply Like
  • varan
    , contributor
    Comments (3521) | Send Message
     
    Author’s reply » YTD Performance 7/22/2013

     

    11.81%

     

    Seems to be OK for a 'balanced' defensive strategy.
    21 Jul 2013, 09:57 PM Reply Like
  • extremebanker
    , contributor
    Comments (1684) | Send Message
     
    Varan:

     

    Your strategy (from what I understand) may be similar to my growth strategy.

     

    I use 15 ETF's. 5 stock, 6 bonds and 4 real assets.

     

    Once per month (every 31 days) I calculate relative strength using ETF Screen. I also calculate relative strength by using total return for last 90 and 182 days. I weight the 90 day calculation at 60% and the 182 day calculation at 40% to determine relative strength for each ETF.

     

    I invest in the top 4 or 5. For the last year it has been stocks only.

     

    I try to overweight small cap and value.
    26 Feb, 03:54 PM Reply Like
  • varan
    , contributor
    Comments (3521) | Send Message
     
    Author’s reply » Thanks Extreme. Multiple variations of the basic theme of starting with a basket of assets with enough of them negatively or only slightly correlated with the others and re-balancing in accord with the evolving market will lead to stable returns, though not necessarily more than the market each and every year.

     

    The strategy summarized here returned 18.9% last year, and has yielded 1% this year so far.

     

    Number of years of losses (2003-todate) 0
    Sharpe 1.07
    Max Drawdown 14.7%

     

    Annualized results so far:
    YTD .95%
    1 Year 16.6%
    3 Years 13.6%
    5 Years 15.3%
    10 Years 11.1%
    26 Feb, 04:22 PM Reply Like
  • extremebanker
    , contributor
    Comments (1684) | Send Message
     
    The reason I started looking at mechanical trading strategies is because of volatility, not performance. I am willing to give up a little performance to avoid big drawdowns. But the results have been very pleasing. I make up so much ground during strong bear markets it makes my performance look very good.

     

    I also limit each ETF to a max of 20% and funds I invest in have to be above their 200 day average.

     

    Look forward to more of your posts regarding strategy.
    26 Feb, 04:43 PM Reply Like
  • drftr
    , contributor
    Comments (115) | Send Message
     
    Varan, Extremebanker, etc.,

     

    Have you guys ever experimented with an adaptive version of Harry Browne's Permanent Portfolio? I mean something like based on momentum / relative strength buying the best performing of VTI-TLT-GLD and if none is above let's say 2% go cash? Do you know anything about the annual performance of such a strategy?

     

    A second question would be if you have done anything similar with levered versions of these ETFs.

     

    And an obvious third and last question if you have played with a setup like in example 1 but with the inverse ETFs included to choose from.

     

    Tnx!

     

    drftr
    5 May, 09:27 PM Reply Like
  • varan
    , contributor
    Comments (3521) | Send Message
     
    Author’s reply » For what it is worth, investing for a quarter in the best of VTI, GLD, TLT, SHY on the basis of prior quarter's performance would have returned 13.5% during 2005-todate, but 2012 (1%) and 2013 (13%) would not have been that great.
    5 May, 10:29 PM Reply Like
  • drftr
    , contributor
    Comments (115) | Send Message
     
    Is there a way I can backtest this for let's say since the early 70s Varan? I know there were no ETFs around back then but are there any websites with monthly prices or so for similar mutual funds and stuff?

     

    drftr
    6 May, 03:55 AM Reply Like
  • varan
    , contributor
    Comments (3521) | Send Message
     
    Author’s reply » I have no idea.
    6 May, 09:48 AM Reply Like
Full index of posts »
Latest Followers

StockTalks

More »

Latest Comments


Instablogs are Seeking Alpha's free blogging platform customized for finance, with instant set up and exposure to millions of readers interested in the financial markets. Publish your own instablog in minutes.