Author Topic: 16.20% CAGR- backtest (1991-2016) after fund fees  (Read 5626 times)

farangster

  • 5 O'Clock Shadow
  • *
  • Posts: 24
16.20% CAGR- backtest (1991-2016) after fund fees
« on: March 14, 2016, 11:42:15 AM »
I used portfoliovisualizer.com for this back-test using the following funds-
VUSTX,EMF,VFINX,VGPMX,DFSCX

I was just trying to find the oldest funds I could find with a bias towards index funds.

timing model- dual momentum
out of market asset- vustx
performance periods- single
timing period- three months
assets to hold- 2
trading frequency- monthly

changing timing period really has a big effect

worst year is -17.21%, max drawdown is -23.72% compared to s&p worst year -37% and max drawdown is -50.95%.  Adding DFCSX adds to cagr.  Any timing period beats S&P significantly during this time period even after fund fees



bacchi

  • Walrus Stache
  • *******
  • Posts: 7056
Re: 16.20% CAGR- backtest (1991-2016) after fund fees
« Reply #1 on: March 14, 2016, 10:42:14 PM »
What's the return compared to the S&P since 2009? From a quick read, it's trailed and is mostly a bear market strategy.

Commission costs will also kill any return unless you're using ETFs at a discount brokerage (IB or Tradestation low).


farangster

  • 5 O'Clock Shadow
  • *
  • Posts: 24
Re: 16.20% CAGR- backtest (1991-2016) after fund fees
« Reply #2 on: March 16, 2016, 10:26:18 PM »
aahhh this is a well known momentum strategy for bull/bear markets.  also if you just use dfa funds with out of market in cash, top 3 asset based on 1 month perf, rebalance monthly- 19.46% cagr over the last 17 years, rebalance qrtly- 18.28% (less trading) if you used vustx instead of cash out of market you get 21.52% qrtly rebalancing over 17 years..........the outperformance is an eyeopener.....fyi trend following has underperformed dismally the last few years in this trendless market.........do I need to say past performance does not guarantee future returns even with a 17 year time period?.......and for the home biased bulls- that goes for the u.s. market the last few years as well!  ; )

bacchi

  • Walrus Stache
  • *******
  • Posts: 7056
Re: 16.20% CAGR- backtest (1991-2016) after fund fees
« Reply #3 on: March 16, 2016, 10:53:42 PM »
aahhh this is a well known momentum strategy for bull/bear markets.

And, what is this well know momentum strategy? A dual momentum relative/absolute comparison strategy signaled to buy VTI in November. That didn't work out so well. It signaled BND in Feb. Another bad idea.

Care to share what the method is or is it mine for only $200?

farangster

  • 5 O'Clock Shadow
  • *
  • Posts: 24
Re: 16.20% CAGR- backtest (1991-2016) after fund fees
« Reply #4 on: March 17, 2016, 08:54:58 AM »
yes this is just a dual momentum but with more than three assets and as i stated initially i was just choosing the oldest funds in different asset classes to get a long timeframe on portfoliovisualizer.........no other reason.  if i was going to invest in this strategy today i would use etf's in a discount brokerage (who doesn't use a discount brokerage?) and the closest comparison to that since 2009 (to answer your question) is VFINX,VUSTX,DFEMX,VGPMX,DFCSX,DFSCX on a quarterly rebalance with 12.02% cagr if the out of asset is cash and almost 18% if the out of asset is VUSTX compared to the S&P 9.58%.  As I stated previously the last few years have been dismal for this strategy, but long term it seems to outperform.....and i am not a daytrader so not interested in short term performance......very easy fiddle with this yourself on portfoliovisualizer.com.

combining relative and absolute momentum for "dual momentum" does not demand just using three assets to choose from to make it work.

folks sure get hostile when you mention ideas that go against what they believe.....wish they would read the entire posts before attacking so i do not have to repeat myself, but i guess discourse is only really a dream!


bacchi

  • Walrus Stache
  • *******
  • Posts: 7056
Re: 16.20% CAGR- backtest (1991-2016) after fund fees
« Reply #5 on: March 17, 2016, 09:05:06 AM »
Tell us the rules. Do just put everything on the relative graph and pick the highest or lowest? And then you compare it to "cash"? Or??? You mentioned holding 2 assets. Is that the top 2 or the top 1 and the bottom 1?

It's hard to "fiddle with this" when the rules are vague.

farangster

  • 5 O'Clock Shadow
  • *
  • Posts: 24
Re: 16.20% CAGR- backtest (1991-2016) after fund fees
« Reply #6 on: March 17, 2016, 09:29:04 AM »
Tell us the rules. Do just put everything on the relative graph and pick the highest or lowest? And then you compare it to "cash"? Or??? You mentioned holding 2 assets. Is that the top 2 or the top 1 and the bottom 1?

It's hard to "fiddle with this" when the rules are vague.

Ok let me hold your hand here to respond to your "vague" comment- go to portfoliovisualizer.com, click on Dual Momentum on the bottom right side of the website, on the next screen put in the tickers (VFINX,VUSTX,DFEMX,VGPMX,DFCSX,DFSCX) under Tickers, change the Timing Period to 1 month, change the Assets to Hold to 3, and the Trading Frequency to quarterly.  Change the start year to 2009 if that is where you want to begin or wherever you want!!

bacchi

  • Walrus Stache
  • *******
  • Posts: 7056
Re: 16.20% CAGR- backtest (1991-2016) after fund fees
« Reply #7 on: March 17, 2016, 09:43:50 AM »
Great, thanks. I'll try it.

MustacheAndaHalf

  • Walrus Stache
  • *******
  • Posts: 6633
Re: 16.20% CAGR- backtest (1991-2016) after fund fees
« Reply #8 on: March 18, 2016, 01:00:38 AM »
...
folks sure get hostile when you mention ideas that go against what they believe.....wish they would read the entire posts before attacking so i do not have to repeat myself, but i guess discourse is only really a dream!
You mean "bacchi" gets hostile, and "bacchi" doesn't read your entire posts.
Your entire thread so far has been with one other person.

workathomedad

  • Stubble
  • **
  • Posts: 189
Re: 16.20% CAGR- backtest (1991-2016) after fund fees
« Reply #9 on: March 18, 2016, 06:36:46 PM »
Can momentum become saturated/overused?

farangster

  • 5 O'Clock Shadow
  • *
  • Posts: 24
Re: 16.20% CAGR- backtest (1991-2016) after fund fees
« Reply #10 on: March 18, 2016, 08:52:40 PM »
Can momentum become saturated/overused?

good question, I don't know but many economists and seasoned investors have said it has existed across the board in most asset classes for over one hundred and the nobel economists Fama and French said something like it was a pervasive long term anomaly.

...........i think it can be similar to stating market cap weighted indexing is oversaturated which is why we see outperformance in fundamental "smart beta" indexing.

 

EarlyStart

  • Stubble
  • **
  • Posts: 115
Re: 16.20% CAGR- backtest (1991-2016) after fund fees
« Reply #11 on: March 20, 2016, 03:21:54 PM »
If you use that entire period, you're probably overfitting the model. It would be a little more robust if you did this, for example:


total years of data- 30
develop strategy based on first 5 years of data, test it on every subsequent 5-year period.

The numbers 30 and 5 are just examples, but you'd probably be better off using this sort of methodology. It's theoretically and statistically more sound.

farangster

  • 5 O'Clock Shadow
  • *
  • Posts: 24
Re: 16.20% CAGR- backtest (1991-2016) after fund fees
« Reply #12 on: March 20, 2016, 04:56:26 PM »
If you use that entire period, you're probably overfitting the model. It would be a little more robust if you did this, for example:


total years of data- 30
develop strategy based on first 5 years of data, test it on every subsequent 5-year period.

The numbers 30 and 5 are just examples, but you'd probably be better off using this sort of methodology. It's theoretically and statistically more sound.
not overfitting that timeframe is simply as far back as I could go with these funds and these funds were not chosen because they had superior past performance to others, which they did not, but because they were the oldest funds I could easily find for their asset classes..... I am sure their were better funds then these just not as old to get a longer timeframe, if you know older funds then these try it out, it's nice to get a cagr after fund fees.  on a side note the market has been a bucking bronco around the 200 sma lately giving a middle finger to trend followers! lol