Testing Dual Momentum in Australia

I wrote this post after getting an email from an Australian reader who had some questions about implementing TADM in Australia. If you have any questions or blog post ideas, just send me an email and I will try answer your questions and share with other readers if appropriate.

Australia presents somewhat of a unique case in the developed market world. Unlike most of the other developed markets—the United States, Japan, the U.K., Germany, France, the Netherlands—Australia does not have a diversified domestic stock market.

The MSCI Australia index, along with it's newer, more popular cousin the S&P/ASX 300 are dominated by financials and materials (the mining sector). These two categories form more than half of the Australian cap-weighted stock market.

The more stable equity categories we see in other developed markets such as consumer staples, consumer discretionary, and industrials together form a little over 15 percent of the market. The newer economy—telecoms and technology—are around 5 percent combined.

The real estate sector of the Australian stock market, at nearly 8% of market capitalization, is larger than the combined size of telecoms, technology, and utilities.

It doesn't take much to figure out that Australia is a natural resource economy with a massive, bloated housing market. Australians are soaked in debt and bound by a wildly swinging currency.

However, although Australians have a relatively narrow local stock market that has tendency to swing wildly with natural resource prices, they may still be much better off investing in a Time Averaged Dual Momentum model.

Investing Considerations for Australians

Given the nature of global markets and easy access thanks to self-directed brokerages, an Australian could simply change their currencies in their trading accounts, buy an American listed ETFs, and follow the standard TADM strategy.

That means buying the U.S. stock index (VTI, ITOT, IVV, etc.), or the Global stock index (VEU, IXUS, etc.), or bonds (AGG, BND, VGIT, IEF, etc.) depending on the TADM signal.

However, the Australian government gives tax preference to investors who hold local stocks. Investors receive a tax credit (franking offset) on dividend income paid by Australian corporations.

It is one of the reasons why Australian corporations pay a high dividend yield by global standard; the yield on the S&P/ASX 300 index is currently over 4 percent.

To take advantage of the tax benefits when investing in Australian exchange listed corporations, this backtest considers whether Australians may be better off investing in a version of Dual Momentum that considers the local index.

It is important to note that Australia's tax system also offers reductions for realized capital gains on investments which are held longer than 12 months.

Unfortunately, many of the signals in my model had holding periods less than 12 months. This means capital gains taxes for an investor following this timing strategy would be a bit higher.

The increased rates for short-term capital gains taxes can be particularly impactful when investing outside of the superannuation system.

Time Averaged Dual Momentum for Australians

In this backtest, I changed the assets for consideration in my Time Averaged Dual Momentum model to consider these three asset classes.

Equities (Relative Momentum Test)

  • MSCI Australia Index
  • MSCI World ex-Australia Index (1970-1987) and MSCI ACWI ex-Australia Index (1988-2017)

Risk-free Rate (Absolute Momentum Test)

  • 90-day Australia Bank Bills Current Yield

All the data was priced in the local currency (the Australian dollar) in attempt to reflect the true returns accrued to an Australian based investor.

It is unlikely that an international investor would see any benefits following this strategy.

As in my other TADM backtests, I did not include the impact of taxes or investment fees on the end investor.

Source: TheRichMoose.com

Time Averaged Dual Momentum (Australia) would have provided the following results for investors.

Compound Annual Return:  +11.72% (48 years)
Largest Annual Gain:  +49.47% (1979)
Largest Annual Drawdown:  -10.72% (1973)
Peak to Trough Drawdown:  -41.28% (1987-1993)

In comparison, an investor buying and holding the MSCI Australia (in AUD) during the same time period would have seen the following results.

Compound Annual Return:  +9.88% (48 years)
Largest Annual Gain:  +70.11% (1983)
Largest Annual Drawdown:  -36.98% (2008)
Peak to Trough Drawdown:  -61.91% (1970-1978)

Summary

This Australia analysis for my TADM model showed some interesting results. There is certainly an advantage in a strict numerical sense for TADM, despite using a poorly diversified, natural resource dominated local market as one of our key signals.

TADM would have provided an investor lower drawdowns on an annual and maximum drawdown basis over the past 48 years.

However, the gains of TADM over buying and holding the Australian index was just 184 basis points annually. This outperformance is not trifling, especially over long time periods.

But once the impact of higher capital gains taxes are calculated on the shorter duration trades the performance advantage of TADM becomes smaller.

We also need to consider if an Australian investor is truly better off choosing this local version of TADM compared with the standard model of investing in U.S. stocks, Global stocks, and bonds.

Some years back, Gogi Grewal posted an analysis on Gary Antonacci's blog looking at foreign investors investing in Gary's Dual Momentum (GEM) model.

When Australians followed Gary's model through converting their currency and buying the appropriate investments on the U.S. stock exchange, the results (measured in AUD) were very positive at +21.9% CAGR over about 41 years.

Although I haven't tried to recreate this backtest method on my similar but unique TADM model measured in Australian dollars, I expect the results to be similar.

The excess gain was mostly due to the Australian dollar declining in value during global recessions, showing a gain to holding U.S. dollar based investments.

While my TADM (Australia) backtest showed some positive results, I believe Australians are going to be better off following a standard Time Averaged Dual Momentum with U.S.-listed ETFs.

If you want to reap some of the tax benefits of holding the local Australian stock index, split your portfolio into two different strategies.

Invest a smaller part of your portfolio in a local buy-and-hold strategy and a larger part in a U.S.-listed TADM strategy.

It may look something like this. (Which would have returned around 14% CAGR since 1970).

Source: TheRichMoose.com

Comments & Questions

All comments are moderated before being posted for public viewing. Please don't send in multiple comments if yours doesn't appear right away. It can take up to 24 hours before comments are posted.

Comments containing links or "trolling" will not be posted. Comments with profane language or those which reveal personal information will be edited by moderator.

How Does Dual Momentum Perform with Leverage

Dual Momentum is a great way for a self-directed investor to take advantage of a simple investment method which has provided higher than index historical returns with lower drawdowns.

My Time Averaged Dual Momentum model uses an average of the 6 month past performance and 12 month past performance in the Dual Momentum analysis. The 48 year backtest shows annual returns averaging more than 16% per year with a max drawdown of 22.47%.

I detail my Time Averaged Dual Momentum model analysis on the Portfolios page of this blog. The page also includes a backtest showing simulated historical results going back to 1970 comparing my Time Averaged Dual Momentum to an equal weight buy-and-hold strategy (66% equities with 33% bonds) over the past 48+ years.

Dual Momentum—as popularized by Gary Antonacci—evaluates three main asset classes. The two equity assets are U.S. stocks compared with International stocks in a relative momentum test.

Following this, the best performing equity asset is compared to the risk-free rate of return—represented by Treasury bills—in an absolute momentum test. The best performing asset is the investor's sole holding for the next month.

Pros and Cons of Leverage

Understanding Problems with Leverage

From what I've gathered reading Gary Antonacci's work, he has typically frowned on the idea of adding leverage to the Dual Momentum model.

His response goes something like this: Dual Momentum still has a lot of short-term volatility, leveraged products (like ETFs) have their own issues which make them less efficient than you think, and the consequent drawdowns of added leverage are likely to be too high for many people.

I don't disagree with most of Gary's perspective. Dual Momentum is designed to be a less risky investment strategy that is easy to follow. Low relative drawdowns means more "stick-with-it-ness" for the average investor.

The value of sticking to a good investment strategy should not be underestimated. Both self-directed investors and professional advisors are notorious for their poor long-term performance purely due to strategy changes and fund turnover.

According to Dalbar studies, the average investor is better off investing 100% of their portfolio in intermediate duration Treasury bonds than trying to invest on their own with various equity strategies.

Aggressive strategies will typically fail in the long run for most investors. Not because the strategy itself failed, but rather because the human executing the strategy failed to keep their emotions in check.

Benefits of Leverage

All this said, I am still a big proponent of proper use of leverage in self-directed investor portfolios. For the smaller group of investors out there who are seeking high overall returns, leverage is the only consistent option.

Other promoted high return strategies are often timing period based, temporary, or inconsistent. Winning stock picking strategies come and go, isolated sector bulls come and go. Even factor-based advantages are not always a great choice.

Proper, risk-adjusted use of leverage is a consistent way of gaining an advantage on the broader markets. This includes use of leveraged ETFs when paired appropriately with safer assets like Treasury bonds, or even short-term high-grade corporate bonds.

Adding Leverage to Dual Momentum

There are two main methods an investor can use to employ leverage with their Dual Momentum strategy.

Using margin to buy standard ETFs is often the better choice in non-registered investment accounts. This eliminates tracking error issues and provides access to a broad range of high volume ETFs.

You can generally deduct the interest expenses on your tax return when you borrow money used to purchase an investment in a non-registered account.

Also, your margin loan interest rates in a good non-registered account are likely to be very competitive at brokerages like Interactive Brokers.

Leveraged ETFs are typically the better choice for accessing leverage in registered accounts. When you borrow in a registered account, you are likely going to pay extremely high interest rates to your broker. In addition, you will not be able to deduct those interest costs from your income on your tax return.

On the U.S. stock exchanges, there are several options for leveraged ETFs for each asset class. Thanks to their growing popularity, there are a number of very high volume ETFs with thin trading spreads.

In my evaluations for adding leverage, I did a monthly readjustment on the leverage. This will vary somewhat from real-world results if you use a daily leveraged ETF, or if you leverage up manually with use of margin. However, the results should be fairly close to both of these methods.

I only applied leverage to the portfolio with the TADM signal was in equity holdings. When the signal is in the bond allocation it rarely makes sense to use leverage as the return differential is not as large.

Adding 1.25x Leverage

I performed three separate leverage tests. This first test was leveraged 1.25x, or adding 25% exposure to the base allocation.

I chose this leverage factor as it is available in the market in ETF format thanks to the Portfolio+ ETFs which are listed on the U.S. exchange. Trading volume is still light in these products, but should be satisfactory for a smaller self-directed investor.

The Portfolio+ products have a current expense ratio of approximately 0.4%. I adjusted the returns down on a monthly basis for each month the portfolio was in equities to reflect the increased cost of this leveraged product compared to standard ETFs.

Source: TheRichMoose.com

TADM with 25% Added Equity Leverage Statistics

Compounded Annual Return:  +19.19% (48 years)
Largest Annual Gain:  +91.66% (1986)
Largest Annual Drawdown:  -12.31% (1973)
Peak to Trough Drawdown:  -29.97% (1973-1975)

Adding 1.5x Leverage

The next leverage test I performed was increasing the equity exposure by 50% on a monthly adjusted basis.

This level of leverage exposure on equities is not currently available in ETF format. If you want to follow a 150% exposure model in a registered account, I would recommend using 2x leveraged ETFs with 75% of your account and leave the remaining account in cash or short-term bonds.

The cost of a 2x leveraged ETF is approximately 0.9%, an increase of about 0.8% over standard ETF products. I adjusted the returns down on a monthly basis for each month the portfolio was in equities to reflect the increased cost of using leverage.

Source: TheRichMoose.com

TADM with 50% Added Equity Leverage Statistics

Compounded Annual Return:  +22.01% (48 years)
Largest Annual Gain:  +115.00% (1986)
Largest Annual Drawdown:  -16.13% (1973)
Peak to Trough Drawdown:  -37.09% (1973-1975)

Adding 2x Leverage

The next leverage test I performed was increasing the equity exposure by 100% on a monthly adjusted basis, effectively doubling the exposure for each signal change.

There are a number of ETFs on the market with 2x leverage on a daily adjusted basis. Even the Canadian exchanges have 2x leveraged products provided by Horizons ETF.

It is also easy to double your exposure with margin loans in a non-registered account. Most margin loans will require 33% equity on the loan. This means a 2x exposure is about as high as you want to go with adequate safety to prevent margin calls.

The cost of a 2x leveraged ETF is approximately 0.9%, an increase of about 0.8% over standard ETF products. I adjusted the returns down on a monthly basis for each month the portfolio was in equities to reflect the increased cost of using leverage.

Source: TheRichMoose.com

TADM with 100% Added Equity Leverage Statistics

Compounded Annual Return:  +28.45% (48 years)
Largest Annual Gain:  +170.41% (1986)
Largest Annual Drawdown:  -23.36% (1973)
Peak to Trough Drawdown:  -49.17% (1973-1975)

Summary

My Time Averaged Dual Momentum model would have performed extremely well when adding leverage to increase the holdings when the signal is in U.S. or International stocks.

The results are simply outstanding, increasing up to a compounded annual return of nearly 30% per year for 48 years at a 2x leveraged equity level. That is a better result than Warren Buffett, John Templeton, and Peter Lynch achieved.

To put that result into some perspective, if you have $100,000 today and you invest in the 2x leveraged model for the next 30 years without any deviation, your portfolio would grow to approximately $183,000,000. (Provided, of course, that future returns are similar to past returns and the costs of leverage stay similar to what they are today.)

However, with increased leverage, Dual Momentum ceases to be a fantastic investment approach for risk-averse individuals. Once you take on a strategy with expected 50% drawdowns, would you be able to stick with the strategy when those drawdowns go to say 60% or higher?

Remember, when you have a model that is historically backtested over an adequate time period, you still need to expect your extreme results to be 25% worse in the future.

Personally I like the 1.25x model for an investor who is looking to add a bit of an edge to their Dual Momentum portfolio. With just 25% leverage, the daily re-balanced ETFs still track very well to the index over long holding periods.

Adding 25% with a margin loan is also very feasible, low risk, and low cost. There is almost no chance of a margin call, plus this leverage amount is easy to maintain on a monthly basis. The leverage is unlikely to run away from you from the time you enter a new position to the time the signal changes to a new position.

Of course a 19+% return compounded over 30 years should not be minimized. That would grow a $100,000 portfolio today into more than $19.3 million dollars. Alternatively, investing $500 a month for 30 years would grow to $9.57 million.

The volatility at 25% leverage is still amazing at less than 20%! Historical drawdowns are under 30%, annual losses of less than 15%, and annual gains up to 91%!

Although a somewhat imperfect measure as the extreme results are tilted to the upside in Dual Momentum, a 25% leveraged TADM model has an impressive Sharpe Ratio of 0.863.

Comments & Questions

All comments are moderated before being posted for public viewing. Please don't send in multiple comments if yours doesn't appear right away. It can take up to 24 hours before comments are posted.

Comments containing links or "trolling" will not be posted. Comments with profane language or those which reveal personal information will be edited by moderator.