Jul 09, 2019
Backtest | Relative Strength
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
Jun 16, 2018
Backtest | Relative Strength | Video
A video showing how to backtest more than 25 securities at a time. The public video below uses the following subscriber-only backtest ETFreplay Relative Strength Backtest - Combine Portfolios
to expand video on screen, click the '4 expanding arrows' icon in the bottom right corner of the video screen
Follow us on
Sep 14, 2016
Relative Strength | Screener
Why does your Relative Strength ranking of ETFs, work better than ranking them using the Sharpe Ratio?
The ETFreplay Relative Strength ranking methodology has the Sharpe Ratio concept at its core but it also reflects some more modern financial modelling methodologies.
So the Sharpe Ratio has volatility in the denominator. The thing about this is that the Sharpe Ratio effectively overrates very low volatility ETFs. In reality, investors value returns more than they do extreme low volatility. For example, a 12% return with 8% volatility is view much more positively than an 8% return with 4% volatility. That move from 8% down to 4% is not nearly as meaningful as the return differential. What investors really want is a solid return with acceptable volatility. Investors can tolerate some level of drawdown with a long-term focus -- just not large drawdowns.
Another thing we did was enable the user to use 2 timeframes for return. The reason is that it is well-accepted that a model can have up to 3 factors as the factors can help each other out. More than 3 factors starts to run into data-mining, which is something we need to be careful of.
Sometimes 1 factor which backtests well over longer time periods can have a rough patch. Another factor can help mitigate the problems and by using 2 return periods, we are not overly reliant on a single return factor.
Hope that helps and let us know if you have any other questions or comments.
See also our FAQ's for common questions: http://www.etfreplay.com/faq.aspx
Follow us on
May 19, 2011
Relative Strength | Video
Exactly 6 months ago we released the Relative Strength Reader application. We review the example done then in the video below:
(original video is on the Tools Tab if didn't see the original).
Follow Us On Twitter:
Feb 25, 2011
Backtest | Relative Strength
Last summer, we added a feature to the portfolio relative strength backtest application. We added a line that showed what the 'provisional picks' would be if the update period had just ended. The provisional picks listed are the exact same as if you ran the ETF screener using the same list of ETFs and same parameters. We added this feature just as a convenience so it wouldn't be necessary to open another tab and check what was strongest now -- rather than at the last update period.
Depending on the list you use, it is of course possible that the 'provisional pick' is not the final pick (hence the term 'provisional'). But what if we just looked up the provisional pick on the next-to-last day of a given month and bought that ETF the next day on the close and held it for the subsequent period? What would the performance look like assuming we did that? Would it be similar?
Let's look at a simple example just to discuss the mechanics of what we mean. We will compare using the 'next-to-last day picks' with the 'last day picks.' The holding periods will be the exact same, we are just reading the picks of the ETF screener with a one day offset. Of course, for many periods the picks (and therefore performance) will be the exact same.
In this example, we will use one of the Ivy Portfolio lists of 5 basic ETFs and a semi-monthly update period. We are choosing the top 1 of 5 and holding it on 2-week intervals. The settings are the exact same except we are checking the box "Invest in next to last day pick(s) near the top of images below.
Using Regular 'Last Day' Picks:
Using Next To Last Day Picks:
We can see that the returns follow similar paths but that there was a difference -- even for this list of just 5 ETFs. In this case, the next-to-last day picks actually performed better -- which means very little in and of itself. We simply are pointing out the mechanics of how it works. Users should test this using their lists across various other settings and draw their own conclusions. Our view is that this doesn't change anything -- if your backtest is well thought-out, then this extra analysis will very likely show a result that is in the same ballpark as the original method.
Final note is that today is Friday, February 25th -- so the next-to-last days picks will be locked in and known after todays close. The final picks will update using Monday's (Feb 28) closing price. In both cases, the backtests will assume the cost-basis for March performance is the closing price as of Feb 28.
Follow us on Twitter:
Jan 11, 2011
Backtest | moving average | Relative Strength
We have added an optional moving average (MA) filter feature to the RS backtest app. With the recently expanded date start and stop functionality, the applications continue to get more versatile.
Combining a long-term moving average within the construct of relative strength has been highly requested and we wanted to discuss one idea when considering whether to use it (note that you can just leave it set to ‘off’ as well).
If you build a relative strength list of say 10 ETFs and you are choosing the top 2, you could protect your portfolio by including 2 bond funds. You don’t need a moving average filter because the bond funds will naturally be the ones with the relative strength when equity markets are dropping. This method actually can get more interesting because you can make better use of more type of ETFs. Rather than just use cash-like bond funds, you might want to extend the potential holdings to an intermediate bond fund like IEF (7-8 year duration) or others. You don't HAVE to restrict yourself to just stocks and cash.
Another way to test is by using a moving average. If you do it this way, then you will inherently be out of ETFs as they go into extended downtrends. You don’t have to proportionally keep X number of bond funds in your list if you do it this way.
But a lot of indexes can go above or below a long-term moving average and still not really be a source of market leadership and enhance your return. Moreover, you may save a lot of money between the time ETFs lose relative strength and the time they actually cross below the moving average. For these reasons, we believe adding relative strength to a MA strategy will generally be more robust. Adding MA to a RS strategy is optional -- and you may find works better or worse than your existing method. Continuous testing leads to better decisions.
Follow us on Twitter:
Dec 08, 2010
Some links to published research on various forms of relative strength concepts. The concept is over a century old. (I am not aware of any studies that involve ETFs directly -- let us know if you know of one).
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1363476 Copy-paste URL into browser.
Abstract from paper #1 above 'Time Series Momentum':
"We document significant "time series momentum" in equity index, currency, commodity,
and bond futures for each of the 58 liquid instruments we consider. We find persistence
in returns for 1 to 12 months that partially reverses over longer horizons, consistent with
sentiment theories of initial under-reaction and delayed over-reaction. A diversified
portfolio of time series momentum strategies across all asset classes delivers substantial
abnormal returns with little exposure to standard asset pricing factors, and performs best
during extreme markets. We show that the returns to time series momentum are closely
linked to the trading activities of speculators and hedgers, where speculators appear to
profit from it at the expense of hedgers. "
Nov 28, 2010
A typical buyside analyst looks at the world in terms of business fundamentals, earnings estimates, quality of management, competitive position, competitive advantage, return on invested capital, P/E multiples and so forth.
You will often hear analysts talk about P/E’s being low relative to growth rates or low vs 10-year bond yields or low vs a group of comparable companies or the overall market etc.
Is the price/earnings (P/E) multiple a good indicator for future performance? One interpretation of this fundamental valuation measure is rather than thinking in terms of the current valuation vs ‘fair value’ --- is simply to observe whether the valuation is expanding or contracting and decide if you believe this will continue.
If a P/E has just gone from 14x to 12x, its cheaper – but is that bullish? How do we know its not headed lower? We should probably test this and see if it works and then create some guidelines for portfolio strategy based on this. Wouldn’t that make sense rather than just blindly believing that a lower multiple is bullish?
Relative strength can actually be thought of as fundamental analysis if you wish to think about it like that. It would not be unusual for a fundamental analyst to say during a rally “the P/E multiple is still too low at 12x next years earnings” -- even though it’s up from a low of 9x and you bought the stock at 10x. But this is not any different than watching relative strength and buying at the same price and looking for further valuation expansion.
Good relative strength analysis captures when a market segment is experiencing valuation expansion. A low P/E is not actually bullish unless its low and then expands higher. This is the crucial aspect – not the absolute level of valuation.
Follow us on Twitter:
Nov 21, 2010
We have received this question a number of times in email and so we wanted to clarify something.
R.S.I. (known as the Relative Strength Index) and the way we at ETFreplay.com discuss ‘Relative Strength’ are not the same thing. We wanted to briefly explain this.
R.S.I. is a technical analysis tool that involves only a single security -- it measures the average amount of up closes vs the average down closes over a given period of time. The most important distinction here is that R.S.I. only looks at the closing values of one security. There is nothing in the calculation of R.S.I. that involves anything but the historical prices of this SINGLE security.
The way ETFreplay uses relative strength has nothing to do with R.S.I. We use relative strength as a way to determine which among multiple market segments is relatively strong.
Many of you are probably familiar with Investors Business Daily. The paper has for a long period of time used a ‘RS Rank’ --- this is more like what we use. Note that IBD certainly did not invent the concept of relative strength -- but they built a database of securities and then ranked everything relative to each other. Institutional-oriented software programs do the same thing. A RS Rank of 90 in IBDs method means that a stock has outperformed 90% of the other securities in their database over a given period of time. High-end institutional software does something similar – except they proceed more mathematically by instead expressing the strength of the security as the distance from the average of a group --- and this is usually stated in terms of the # of standard deviations away from the mean (think z-score).
This is all similar to what we have done – except we do it with ETFs and then allow you to backtest it yourself rather than just saying ‘you should buy relative strength because it works.’ We also allow the user to define relative strength themselves – using easy browser controls like drop-down menus and text boxes. So for example you could simulate IBD’s method by using 12-month performance and ranking ETFs in a given universe like this:
Then you could go the ETFreplay.com backtesting module and see how 12-month relative strength has been holding up over the past 10+ years and see what kind of drawdowns its had.
This is all a research process – its just that we are performing research that comes in a very practical form. We aren’t researching stocks, we are researching strategies --- strategies based on baskets of stocks. Backtesting is not the only thing that matters --- but its pretty darn good information for you to factor into your decision-making process. Without some historical testing, you could easily go a lifetime of doing things that you thought worked – but actually don’t -- and never really did.
Follow us on Twitter: