ETF Mean Regression and ETF portfolio backtesting

Aug 27, 2019 in Backtest | Video

A video to demonstrate mean-regression on and some concepts related specifically to ETF mean-regression.


 to expand video on screen, click the '4 expanding arrows' icon in the bottom right corner of the video screen


Follow us on Follow etfreplay on Twitter


Diversification comes in various forms

Jul 09, 2019 in Backtest

Maintaining a well diversified portfolio is a time-tested way to protect against going all-in on what turns out to be a terrible investment.  Diversification can also be employed at the strategy level for the same reason.  An example of this is the core-satellite framework, where a rebalanced core portfolio is mixed with different strategies that focus on Relative Strength and Moving Average trend following etc. 

It is also possible to diversify across different versions of a single strategy, to reduce the risk of parameter choice misfortune.  For example, rather than relying solely on 12-month returns, for instance, the backtest below equal weights 4 variants of the same model: a 6-month version, 8-month, 10-month and one using 12-month returns all on the same ETFs: EFA, IEF and VTI (the constituents of BNCH).


click image for full size version

Just as a well diversified portfolio means that at least some part of it will always be a drag, a composite made up of different model variants will always underperform the best version of the strategy....but it also avoids being exclusively in the worst.


Follow us on Follow etfreplay on Twitter


Emerging Markets Lookback Period Fund Backtest Results Array

May 15, 2019 in Backtest | Emerging Markets

One variable to analyze more deeply is the 'Lookback Period' -- this is the period through which you filter out noise.  You obviously don't chase whatever performed best today vs yesterday (a 1-day lookback)... Nor should you chase whatever performed best over the past 5 years (laggard).   Academia has demonstrated in numerous research papers going back many, many decades over many different asset classes that the 3 to 12 month range is the value-added focus zone.  

Set-Up:  We test Emerging Markets over a 16-year period using both monthly lookbacks {3, 4, 5, 6, 7, 8, 9, 10, 11 & 12 months} --- and then also test weekly lookbacks {13, 17, 22, 26, 30, 35, 39, 42, 48 & 52} weeks.   The monthly tests use last day of month to execute rotation, when there is rotation.   The weekly tests use the last trading day of the week (ie, Friday close).    We use zero interest security (XZERO) as our holding when not invested (this is unrealistically conservative but it allows us to compare [3] month with [12] month lookbacks without the minor complication of changes in fixed-income total returns when out of the market).  

Results:  Generally speaking, emerging markets are better examined through a 4-5-month (17-22wk) range in terms of lookback period.  10-11 month were profitable but far less in total return over this period than the other time periods.    Weekly and monthly time periods show similar results showing that variation will occur – we are only looking for larger tendencies here.   Secondarily, 3 & 6 month (13-wk and 26-wk results) proved better than 9 & 12 month.   

VEIEX_16_Year_Backtest_Array_Slide.pdf (129.62 kb)

Follow us on Follow etfreplay on Twitter


Strategy Diversification: Combine a core allocation with regime based portfolio switching

May 07, 2019 in Regime Change

Back in 2010 we created our first multiple strategy module, the Advanced Relative Strength backtest, allowing subscribers to combine together different models into an overall portfolio.  To illustrate the backtest, we produced a simple example that employed two sub-strategies; a basic US equity model (MDY, IWM, SPY and QQQ) and an international model using smaller developed country funds (EWA, EWC, EWH and EWS).

The example below uses the same ETFs as that original illustration, but this time, rather than running each model concurrently, we have employed the SPY / EFA ratio moving average as a regime switch to dynamically alternate between the two portfolios.  When the SPY / EFA ratio is trending upwards (i.e. above its MA), the backtest invests in the US equity portfolio.  When the opposite is true, it switches to the International stock portfolio.  This regime approach is then mixed with a solid fixed income core portfolio (IEF and LQD) to form an annually rebalanced 60-40 strategy.


The Core-Regime Portfolios backtest is available to pro subscription members


Follow us on Follow etfreplay on Twitter


Regime Change Backtest Example

Apr 24, 2019 in Regime Change

Regime Change is used in finance to describe when a condition changes.   IF [condition1] is met,  THEN invest in [X]...   ELSE invest in [Y].   ETFs allow us to easily test conditions which are defined not by some calculation you've created to simulate an index,  these are publicly traded securities with real money invested in them.  There is no ambiguity as to the rules when you use real-world securities as is so often the case with non-financial regime tests.  

Here is a simple example to get the hang of it, is the NASDAQ-100 going up?  If it is, buy it.  If it isn't, invest in a different type of ETF.   In this example, the different type of ETF is defined by the QUALITY FACTOR.   Quality stocks are those with strong balance sheets, lower earnings variability & higher Return On Equity -- as ranked by indexing firm MSCI.   QUAL actually owned real stocks on each day with real money, we aren't subjectively now determining what should be classified as quality and what shouldn't.

What does the performance report look like for this idea?  See below for summary version of an backtest report  (statistical analysis excluded in image below).

Then try other ideas.   All of your ideas don't have to work for you to be very successful at this.   Indeed, this strategy has underfperformed its benchmark 46% of the time in last 5 years (as measured by relative performance in each calendar month).  Yet the outperformance over time has been good.


Follow us on Follow etfreplay on Twitter


Uptrending Ratio Indicates Relative Strength ETF Backtests QUANTIFY It

Apr 23, 2019 in Backtest | Ratio


An uptrend is a series of higher highs and higher lows.   Using a ratio between 2 securities shows which is relatively stronger.   A Relative Strength analysis can quantify which security within a list of MORE THAN 2 securities is strongest.  

So let's look at one current situation.    Emerging markets have shown good relative strength on shorter-term basis.  If this continues then a higher low and higher highs situation could develop (vs SP 500).   That said, SPY has continued to be strong -- both Emerging markets AND US Stocks have been strong this year.   It actually hasn't mattered which you've owned --- so even if you were wrong on thinking a ratio would go up/down, you still made good money either way.    This won't always be the case though.   


Follow us on Follow etfreplay on Twitter


Channel Backtest Example -- China A-Shares ETF ASHR

Apr 05, 2019 in Backtest | Channel

Channels are a good, simple supplement that offer an ABSOLUTE look and can be used in conjunction with other RELATIVE studies. Wider channels give your trade room to work. Tighter channels will cause some whipsaw losses. If you are bullish on an ETF based on a range of factors, then running a skewed channel might be a good idea --- ie, run the exit (Sell channel) at 0% but a buy at just 60%... This allows you to get in quickly while still offering room for the investment to work. This study uses a simple 67% / 33% buy/sell trigger with a ~6 month lookback (26-weeks means you will trade usually on a Friday -- if holiday then Thursday). Entries and exits only occur on the close of the last day of the week (Fridays) allowing for easy monitoring. Note that this look has trades that have lasted a while. This is because the sell rules will allow a fair bit of movement before exiting.

Finally, because a channel uses a percentage,  it may be easier to see the trend in the ETF than othewise.   An uptrend is defined by higher lows and higher highs.    A downtrend is defined by lower highs and lower lows.   The channel is a another tool to have to see and understand what is happening in the market.


Follow us on Follow etfreplay on Twitter


Introducing Dashboards - A Way To Help Organize Workflow In Research & ETF Portfolio Backtesting

Mar 05, 2019 in Video | Dashboards

A video to demonstrate one way to use ETFreplay Dashboards -- a new feature on the site to help you monitor your various strategies. We view this as nicely complementary and a useful dimension to add to the site.


 to expand video on screen, click the '4 expanding arrows' icon in the bottom right corner of the video screen


Follow us on Follow etfreplay on Twitter


New feature: Dashboards

Feb 14, 2019 in Dashboards

We have added Dashboards as a new feature to the website so that subscribers can easily view multiple models and/or markets in one place.

Dashboards can be used in numerous different ways, including:

  • to view market activity from various angles with a mixture of several different models
  • to look at one model applied to several different ETFs / markets - see MGI Report of Jan 17th
  • to monitor different variations of a single model

etc etc.

Most importantly, you have the choice of how to set up your own personal dashboard(s).

To create a dashboard:

  1. From 'My Account' in the top menu, choose 'My Dashboards'
  2. Click 'Add Dashboard' button in the top right corner, enter a name and click save
  3. A dashboard with 6 empty windows will now appear
  4. Click the 'Add Item' button in top right corner of one of the empty windows
  5. Choose between:
    1. 'Screener'
    2. 'Ranks'
    3. 'Ratio Chart'
    4. 'MA Chart'
    5. 'MA List'
  6. Enter the rest of the required details, click 'Save' and the desired chart or table will appear in the window.
  7. Repeat steps 4, 5 and 6 for the other empty windows, up to a total of 6 per dashboard
  8. Items can be edited / removed by clicking 'Edit Item' in the top right corner of the required window
  9. The 'through' date can be changed via the calendar control in the top left corner

Subscribers can create up to 5 dashboards. Pro Subscribers can have up to 20.



Follow us on Follow etfreplay on Twitter


Jan 31, 2019 ETFreplay Market-Generated Information (MGI) Report ETF Backtest

Jan 31, 2019 in Ratio | Regime Change

 Jan 31, 2019  ETFreplay Market-Generated Information (MGI) Report

Focus on using TIPS index as a backtest parameter for Energy Stocks (XLE)

(Click on Image For Link to PDF)


Follow us on Follow etfreplay on Twitter


Follow ETFreplay