A Gentle Guide to Global Tactical Asset Allocation

by Corey Hoffstein, Newfound Research

  • Two questions we frequently receive are: “what is global tactical asset allocation?” and “what are style premia (factors)?”
  • In this commentary, we aim to provide a very high-level answer to those questions, incorporating as little math or financial theory as possible and avoiding nuanced discussion. This is not meant as a practitioner’s guide, but simply a very basic introduction.
  • At the end of this commentary, our goal is for anyone who reads it to walk away with a strong intuition about global tactical asset allocation and to demystify the word “style premia.”

What is Global Tactical Asset Allocation (“GTAA”)?

Most invest to put their money to work.  By lending (with bonds) or investing (with stocks), investors seek to grow their money that might otherwise sit, largely stagnant, in a savings account.

Underlying the idea of investing is the fundamental notion that if we take on risk, we should expect a positive reward.  When we buy bonds, we are lending our money to a company.  We expect to earn a return on this loan because there is a risk the company might not pay us back.  Similarly, when we invest in a stock, we are buying a piece of a company.  We expect a return because there is a risk the company might go bankrupt.

After all, if we did not expect a positive reward, why would we risk our hard-earned capital?

Traditional active investors believe that through disciplined security analysis, it is possible to identify opportunities for outsized return in the market place.  To illustrate this concept, Benjamin Graham, and his disciple Warren Buffett, popularized the character Mr. Market.  Mr. Market is your fictitious business partner who frequently offers to buy or sell his shares of the business.

Mr. Market also has a somewhat manic-depressive personality.  Most days, he offers to buy and sell from you at a fair value.  On occasion, however, he becomes wildly optimistic or pessimistic, offering to buy and sell at extremely inflated and discounted prices.  Prudent investors should capitalize on this opportunity.

Graham and Buffett’s allegory articulates a view that the market is frequently, but not always, rational.  Global tactical asset allocators agree with this view, but differ in their implementation.

While Graham and Buffett may establish a view of how much Coca-Cola is worth and wait for the stock to trade at a deep discount, global tactical asset allocators establish views at higher levels of the investment hierarchy.

What is the investment hierarchy?  Broadly speaking, the hierarchy is:

  • Asset Class: The type of security (e.g. stock, bond, real estate, commodity, currency, etc.)
  • Geography: The region or location (e.g. United States, Germany, Japan, China, etc.)
  • Sector / Industry: The segment of the economy (e.g. Financials, Technology, etc.)
  • Company: The firm issuing the security (e.g. Apple, Coca-Cola, etc.)

As an example, a share of “KO” is a stock in the United States, falling in the Consumer Staples sector, issued by Coca-Cola.

Global tactical asset allocators generally make investment calls at the asset class, geographic, and sector / industry level.  Some will then go on to use these views as a guide to where they should do further security research, while others may simply buy a broad basket of securities to express their view.

Let’s look at some examples of how opportunities at these levels of the hierarchy may manifest.

Asset Class Level Opportunities

As we said at the beginning, we invest because we expect to earn a return.  The riskier the investment, the more we expect to be rewarded.  While we expect bonds to return more than cash sitting in our bank account, we similarly expect stocks to outperform bonds.

What we expect and what actually happens, however, are not always the same.  Stocks are, generally, riskier than bonds, and can go through periods of significant loss.  A recent example is the Global Financial Crisis, which caused U.S. stocks to lose more than 36% of their value in 2008.  Bonds, however, held up nicely in this period, returning almost 8%.  On the other hand, as the outlook cleared in 2009, stocks rallied 26% while bonds only returned just under 3%.

A global tactical allocator may look to exploit this performance dispersion, trying to shift their allocation towards bonds in environments like 2008 and towards stocks in environments like 2009.

Source: CSI.  Calculations by Newfound Research.

Geographic Region Opportunities

Similar opportunities may exist within a given asset class.

After World War II, Japan entered a period of rapid economic growth that would later become known as the Japanese economic miracle.  By the 1960s, it emerged as the second largest economy in the world, eclipsed only by the United States.

Strong economic performance led to strong performance of Japanese equities.  From 1969 to 1989, the MSCI Japan index returned 5598%.  As a share of the global investible equity market, it grew from just north of 10% in 1975 to 40% in 1988.

Over the next decade, Japan’s economy stagnated.  The MSCI Japan index returned -8% from 1989 to 1999, and Japan’s share of the global investible equity market shrank back to about 10%.  In fact, it was only in May 2017 that the MSCI Japan index eclipsed its prior high in February 1989, some 28 years earlier.

Global tactical asset allocators may seek to exploit this type of opportunity by increasing their exposure to Japanese equities during periods of aggressive economic growth and reducing, if not outright eliminating, their exposure when they feel Mr. Market has perhaps become a bit too optimistic about the outlook.

Below we plot the percentage each country has made up of the global investible equity market.  The United States is identified as the black region with white dots, while Japan is the white region with black dots.

Source: The World Bank.

Note that we can see a similar effect in U.S. equities from 1989 to 2002, during the tech-boom that drove up U.S. equity prices.  Which brings us to our next level of the hierarchy.

Sector Level Opportunities

Examples that may be more front-of-mind for U.S. investors are the dot-com bubble of the late 1990s and the Great Financial Crisis from 2007-2009.  The massive impact of these events on U.S. equity returns was largely due to the high degree to which U.S. equity markets were exposed to the companies involved.

Below we break down the composition of the S&P 500 into sectors.  The Technology sector is identified as the black region with white dots and the Financial sector is the white region with black dots.

Source: CSI.  Calculations by Newfound Research.

In February 2000, the Technology sector represented over 45% of the S&P 500.  By comparison, the Utilities and Materials sectors each represented less than 2%.  Similarly, in December 2007, the Financials sector represented over 30%.

These bubbles and subsequent collapses represented opportunities – both for earning outsized returns and avoiding losses – for global tactical asset allocators, who have the flexibility to significantly shift their sector exposures.

Understanding Investment Styles

While the prior examples largely focused on extreme market dislocations and bubbles that were obvious only in hindsight, these are not the only types of opportunities that a global tactical asset allocator might seek.  Views may be much tamer, such as expecting U.S. equity returns to be lower than they have historically been, or expecting emerging market sovereign debt to outperform U.S. Treasuries by a rate that exceeds the excess risk.

What is important is understanding how a global tactical asset allocator comes to these views.

At Newfound Research, we explicitly focus on systematic – or rules-based ­– approaches to investing.

There are five common styles of systematic investing: value, momentum, carry, defensive, and trend.  While each investment manager may have their own unique approach to implementing these styles, they generally follow the same overarching concept.  Painting with broad strokes, the styles can be generally described as:

StyleWhat is It?Why Do We Think It Works?ValueIdentify investments that are trading at a steep discount or premium to their intrinsic value.Reversion to the mean; discomfort required to disagree with the market.MomentumIdentify investments that have been recently out- or underperforming their peers.Behavioral biases exhibited by investors such as herding.CarryIdentify investments that offer a higher return with the assumption prices do not change.Compensation for higher risk, supply and demand imbalances, and structural rigidities.DefensiveIdentify investments that are “safer.”Behavioral biases that lead investors to eschew leverage and seek lottery-like opportunities.TrendIdentify investments that are appreciating in value.Behavioral biases exhibited by investors such as herding.

Empirical and academic evidence suggests that each of these investment styles generates a statistically robust “premium” (extra return) that investors can earn over time.

These are, by no means, the only way global tactical asset allocation decisions can be made.  Many of the most famous global tactical calls have been made on a more discretionary basis that exploit a unique and idiosyncratic situation.

For example, in the early 1990s, George Soros’ Quantum Fund famously “broke” the British pound.  At the time, Soros saw England’s submission to the European Exchange Rate Mechanism (“ERM”) as a fundamentally unsound position due to inflation differences between England and Germany.  Betting against the pound, the Quantum Fund reportedly made $1 billion when England withdrew from the ERM.

While such tales of speculative heroics may make for good headlines, the idiosyncratic nature of such events makes it very difficult to design a consistent investment plan around.  We believe the systematic approach of the primary investment styles allows for much more transparent setting of performance expectations.

Identifying Investment Opportunities with Styles

There are two primary types of calls made by global tactical asset allocators: absolute and relative calls.  Absolute calls are those made about the outlook of an investment opportunity irrespective of other available investments.  Relative calls, on the other hand, explicitly compare and contrast investment opportunities against one another.

Absolute Calls

Earlier we said that we expect to earn a return on our investment because we are taking risk.  We also said that riskier investments should demand a higher expected reward.  Here is a question: should the expected reward necessarily be constant over time?

For example, it is generally expected that stocks should outperform cash in the long run.  The excess return that stocks are expected to earn is known as the equity risk premium.  One question we might ask is, “do we expect this premium to be constant over time?”  As an example, do we believe stocks will earn the same premium over cash at the top of the tech-bubble in 1999 as the bottom of the Great Financial Crisis in 2009?

Global tactical managers that make absolute calls believe the answer is “no,” and seek to invest when the return opportunity is rich and avoid when it is thin (or outright negative).

We won’t sugar coat it: absolute calls are market timing.  To call market timing a controversial strategy is an understatement.  We believe, however, that there is a significant difference between the ad hoc prediction of market tops and bottoms, which has little supporting evidence, and the systematic implementation of investment styles.  Specifically, we believe evidence supports the use of value, defensive, and trend styles for making absolute calls.


A value-based approach to market timing is effectively saying, “I believe this investment is worth X.  I am absolutely willing to buy it at Y and will never touch it at Z.”  The idea behind value is that Mr. Market is behaving in a manic-depressive manner and that at Y the investment is very undervalued and at Z the investment is very overvalued.  The art and science is in determining exactly what X, Y, and Z are.

For a stock market index like the S&P 500, an aggregate fundamental metric is normally chosen – for example, the aggregate earnings of all the stocks in the index – and compared against the price.  In a vacuum, a metric like price-to-earnings does not tell us much.  However, if we compare it against historic levels, we may be able to glean some information about the aggregate investor’s risk appetite.

Below we plot the CAPE Ratio, which is the price-to-earnings ratio for the S&P 500 smoothed over a 10-year period.

Source: Robert Shiller.

One way of interpreting this number is that if the earnings of the S&P 500 did not grow, and 100% of earnings were paid out as dividends, then the number is how many years it would take for you to earn back 100% of your investment.  In 1982, this number was south of 7; in 1999, it peaked over 44.

High CAPE ratios are, then, often interpreted as “overvalued” and low CAPE ratios as “undervalued,” with “fair value” for CAPE often being defined as a multi-decade average.  We can use this idea to create a value-based timing strategy, where we invest in the market when it is cheap and divest when it is expensive.

Source: Robert Shiller; Kenneth French.  Calculations by Newfound Research.  The Value Timing Strategy uses an expanding window process to calculate a Z-Score for the Shiller CAPE ratio.  Z-Score values are clamped between -3 and 3.  Z-Scores are then transformed into weights, with -3 representing a 50% weight and 3 representing a 150% weight to the U.S. equity market.  The portfolio is rebalanced monthly.  A 60-month overlapping portfolio process is employed.  Capital is assumed to be borrowed and lent at the risk-free rate.  Strategy returns assume the reinvestment of all distributions.  Strategy returns are net of all fees, including trading costs and taxes.  Returns are purely hypothetical and backtested.  Past performance is not an indication of future results. 

While our above example focused exclusively on the U.S. equity market, a similar concept can be applied to other asset classes (e.g. “are bonds expensive or cheap?”), geographies (e.g. “is the Brazilian Real expensive or cheap?”), and even sectors (e.g. “are high yield bonds of energy companies expensive or cheap?”).

Determining fair value is no easy task.  Empirical evidence suggests that value-based timing may be more appropriate in the extremes (e.g. never-before seen under- or overvalued levels) than for values in the middle.  For example, even after the dot-com bubble burst (bottoming in early 2003), U.S. equities have largely sat above their long-term average price-to-earnings ratio.  During this “overvalued” period, U.S. equities have returned over 300%.

Betting against the market, and waiting for mean reversion, can be painful.  To quote the famous economist John Maynard Keynes, “the market can remain irrational longer than you can stay solvent.”


Momentum is strictly a relative performance investment style.  As an absolute style, it is called trend, and is discussed below.


Carry is the return earned from simply holding an asset, assuming zero price appreciation.  For example, if we buy a bond with a yield-to-maturity of 3%, we earn a 3% yield every year, regardless of what interest rates do.  That is the carry.  Not all assets have a positive carry, however: if I buy a barrel of oil, there is a physical cost to storing that oil.  If I sell that oil 6 months later for the same price I bought it at, I will have lost money: the cost of carrying the oil.

Market timing with carry is predicated on the idea of determining whether the yield being earned on an investment is sufficient to remain invested or not.  For equities, where carry is measured as the dividend yield, this strategy is a cousin of value.  Assuming no change in fundamentals, when prices fall, both yields and cheapness measures richen.

For other asset classes, however, it can be an entirely independent measure.  Consider, for example, U.S. Treasuries.  When we buy a U.S. Treasury, we have the choice as to the maturity of the bond: in essence, how long do we want to lend our money for?  In theory, lending our money for a longer period of time is riskier, and therefore should demand a higher return.  Hence, the emergence of the term premium: the excess return of longer-dated bonds over shorter-dated ones.

Measuring this premium, ex-ante, is not obvious.  However, one estimate is to look at the difference between the yield of a 5-year U.S. Treasury versus a 2-year U.S. Treasury (often called the “slope” of the yield curve).  The difference between the two will capture the expected premium that can be earned by lending our money for an extra three years.  Investors may choose to take the extra risk when the premium is high and avoid the risk when the premium is low, essentially timing their interest rate exposure risk. 


In its interpretation as a style, defensive is a rather broad category.  The common characteristic is that measures try to establish how “safe” an investment is, from both a qualitative and quantitative stand-point.

For market timing, such an approach could involve volatility targeting.  Volatility is a statistical measurement that captures the magnitude of dispersion in returns.  Broadly, low volatility means small movements in price, while high volatility means the potential for big jumps.

While different asset classes have different levels of volatility (e.g. stocks typically have higher volatility than bonds), volatility for an asset class can also change over time.  A volatility targeting approach, then, is a strategy that tries to dynamically allocate in such a way that the volatility profile of an asset class remains constant over time.  This is achieved by holding cash when volatility is too high, and using leverage when volatility is too low.

Source: Kenneth French.  Calculations by Newfound Research.  Realized Annualized Volatility is calculated as a 12-month exponentially weighted moving average.

A volatility targeting approach also seeks to exploit the empirical evidence that volatility spikes during significant market sell-offs, potentially allowing the strategy to lose less in significant bear markets.

Source: Kenneth French.  Calculations by Newfound Research.  The Defensive Timing Strategy uses a rolling 12-month exponentially weighted volatility calculation.  Position size is dictated by a target volatility of 16% (as estimated using the trailing realized volatility measure).  Market exposure is capped between 50% and 150%.  The portfolio is rebalanced monthly.  Capital is assumed to be borrowed and lent at the risk-free rate.  Strategy returns assume the reinvestment of all distributions.  Strategy returns are net of all fees, including trading costs and taxes.  Returns are purely hypothetical and backtested.  Past performance is not an indication of future results.

Again, while we have demonstrated this approach using U.S. equities, there is no reason it cannot be applied elsewhere in the investment hierarchy.


In a trend-following approach, a global tactical asset allocator compares the recent return of an investment against a hurdle rate (often cash).  When that return is positive, a positive trend is assumed; when the return is negative, a negative trend is assumed.  These trends are expected to continue to persist over short-term horizons, giving the tactical investor the ability to forecast what direction the investment might move.

Exactly why trends have historically existed is still largely up for debate.  Academic literature has most recently focused on behavioral arguments.  Specifically, when the fundamental value of an investment changes, investors exhibit biases that cause them – and therefore the price – to underreact to this new information.  Once the trend begins, however, other investment biases emerge, causing the trend to overshoot fair value.

Below we apply a simple trend-following strategy to U.S. equities, investing during periods of positive trends and divesting during negative trend periods.

Source: Kenneth French.  Calculations by Newfound Research.  The Trend Timing Strategy uses a rolling 12-month return to determine the trend.  When the return is positive, a positive trend is assumed, and a 150% position is held.  When the return is negative, a negative trend is assumed, and a 50% position is held.  The portfolio is rebalanced monthly.  Capital is assumed to be borrowed and lent at the risk-free rate.  Strategy returns assume the reinvestment of all distributions.  Strategy returns are net of all fees, including trading costs and taxes.  Returns are purely hypothetical and backtested.  Past performance is not an indication of future results.

Perhaps the most well-known approach to trend-based market timing is a category of investment known as Managed Futures.  Managed Futures strategies apply trend-following across a broad array of global asset classes, including equities, bonds, currencies, and commodities.  The breadth of investment opportunity, as well as the ability to go both long and short to profit from both positive and negative trends, has allowed Managed Futures strategies to historically serve as a valuable diversifier in periods of crisis.

Expectations with Market Timing

There is a reason that market timing is controversial: there is little empirical evidence that investors have been able to do it with any sort of consistency. Managed Futures – a multi-asset, diversified approach to market timing – stands alone as one of the few approaches that has consistently exhibited strong excess risk-adjusted returns.

What the above backtests hide is the extreme discomfort these strategies can generate when they underperform.  Below we plot the relative performance of all three strategies versus U.S. equities.  When the graphs are going up, the strategy is out-performing equities; when the graphs are going down, U.S. equities are out-performing the strategies.

Source: Robert Shiller; Kenneth French.  Calculations by Newfound Research.  Please see above disclosures for the Value Timing Strategy.

Source: Kenneth French.  Calculations by Newfound Research.  Please see above disclosures for the Defensive Timing Strategy.

Source: Kenneth French.  Calculations by Newfound Research.  Please see above disclosures for the Trend Timing Strategy.

These strategies all share two things in common.  First, they all out-performed their benchmark over time.  Second, they all experienced decades of relative underperformance: Value from 1990 through 2000, Defensive from 1969 to 1992, and Trend from 1956 to 1967.  They also all exhibited periods of sharp and sudden relative underperformance: Value in from 2002 to 2004, Defensive in 1998 to 1999, and Trend in 2015 to 2016.

In other words: it was a painful ride.  Most investors likely would have abandoned these strategies somewhere along the way.  In fact, we would argue that for these types of strategies to work, they have to be that painful.  More on that later.

Relative Performance Calls

As can be expected from the name, relative performance calls evaluate the attractiveness of an investment not based upon its stand-alone merits, but in comparison to other investments.  For example, below we plot the growth of $10,000 invested in funds tracking the S&P 500 and the MSCI United Kingdom indices.  We can see that both indices lost money during the dot-com fallout, rallied in the recovery, lost value again in the 2008 credit crisis, and subsequently grew thereafter.

Source: CSI Analytics.

The high degree of correlation in these returns tells us that there is likely a common underlying risk factor.  In this case, likely some exposure to global economic growth.  There is another way to look at this performance, however: how the two indices performed relative to one another over time.  Below, we plot the relative performance of the S&P 500 versus the United Kingdom.

Source: CSI Analytics.  Calculations by Newfound Research.

Through this lens, we see that from 2000 through 2008, the S&P 500 largely lagged the United Kingdom.  After 2008, however, it out-performed.  Had we perfectly timed this transition – investing first in the United Kingdom and then in the S&P 500 – we would have earned an extra $8,200 on our initial $10,000 investment.

Source: CSI Analytics.  Calculations by Newfound Research. 

The benefit of relative value calls is that their value-add can be independent of the other risks taken in the portfolio.  For example, a moderate investor may allocate 50% of his capital to stocks and 50% of their capital to bonds.  Relative trades of where to allocate within the stock sleeve, however, may create excess return without dramatically shifting the overall risk profile of the portfolio.


A value-based approach to relative performance calls relies upon identifying exposures that are trading cheaper – with respect to an estimate of fair value – than their peers.  This is the traditional Benjamin Graham and Warren Buffett approach applied at higher levels of the investment hierarchy.  For example, instead of investing in cheap stocks, a global tactical asset allocator might invest in the broad equity indices of cheap countries.

As a concrete example, below we apply a value-based strategy to the ten primary U.S. sectors.  Each month, the strategy ranks the sectors based upon their valuation.  After ranking the sectors, they are paired into 5 groups, representing the cheapest to the most expensive.

The difference between a relative approach to value and an absolute approach to value is that the relative approach does not care what the level of valuation is for an investment, simply what its relative valuation is.  For example, it may be the case that when evaluated on their own merit, all the sectors would be considered overvalued.  In this case, while an absolute approach may choose to not invest, the relative approach would look to invest in the sectors that are the least overvalued.

Source: Kenneth French.  Calculations by Newfound Research.  The value score for each sector is measured as a z-score of current yield using the prior 60-months of data.  The strategy is rebalanced monthly using a 60-month overlapping portfolio approach.  Strategy returns assume the reinvestment of all distributions.  Strategy returns are net of all fees, including trading costs and taxes.  Returns are purely hypothetical and backtested.  Past performance is not an indication of future results.

While the portfolio that held the most expensive sectors returned 9.88% per year – turning $10,000 into $18.9 million after 80 years –the cheapest portfolio returned 12.61%.  The extra 2.73% annualized may not seem like much, but over the 80-year period, $10,000 turned into an astounding $135 million.

While all of the cohorts exhibit a high degree of correlation to one another – because of their shared exposure to broad U.S. equity risk – the outperformance of the cheapest versus the most expensive cohort follows its own path.  This relative performance is plotted in the graph below.

Much like our examples with market timing, we can see that the value approach is not infallible: it too suffers periods of significant and prolonged underperformance.  Most notably, this strategy has largely under-performed since 2004.  What is valuable, however, is not just the opportunity for outperformance, but that this outperformance represents a potentially independent return stream from traditional market factors like stock and bond returns.

Source: Kenneth French.  Calculations by Newfound Research.

Exactly why value-based strategies have historically worked is still highly debated.  Some argue that securities trading at a discount to their intrinsic value do so because they have a riskier outlook.  Purchasing such securities requires bearing more risk, and therefore, over the long run, you should be compensated for it.  Others argue that there are structural and behavioral biases – like the herding behavior of investors – that cause relative value opportunities to emerge.

The horizon over which mean reversion takes place can be years.  Being a contrarian for such a long period can be an emotionally painful experience – but one that we are hopefully rewarded handsomely for.


In physics, momentum is the product of mass and velocity.  For example, a heavy truck moving at high speed has large momentum.  To stop the truck, we must apply either a large or prolonged force against it.  Momentum investors invoke a similar concept: they expect investments that have recently outperformed their peers to continue outperforming and those that have underperformed to continue underperforming.

That recent relative performance is predictive of future relative performance is one of the strongest and most persistent anomalies found in financial literature.  There are several theories as to why relative performance trends emerge and persist, including: behavioral biases exhibited by other investors (e.g. herding), structural limits of investors, and limited attention capacity.

Whatever the reason, while value investing bets against the herd, momentum explicitly invests with it.  Momentum investors wait for relative trends to emerge and then try to ride them for as long as they persist.

Consider the original example of relative performance calls: the S&P 500 versus the United Kingdom.  A momentum strategy would look at recent performance and invest in the index that has the highest total return.  We construct such a strategy below.

Source: CSI Analytics.  Calculations by Newfound Research.  Each month, the SPDR S&P 500 ETF (“SPY”) and the iShares MSCI UK ETF (“EWU”) are ranked on their trailing 12-1 month total return.  The ETF with the highest total return is held for the next month.  Strategy returns assume the reinvestment of all distributions.  Strategy returns are net of all fees, including trading costs and taxes.  Returns are purely hypothetical and backtested.  Past performance is not an indication of future results.

You may notice that this strategy ends up with a similar total return to the S&P 500 over the period.  At first glance, this may seem like an underwhelming result.  However, while this is evident ex-post, ex-ante we would not have known which of the two indices would have performed best during this period.  Therefore, we can say that this strategy was able to deliver to us the same result we would have achieved had we had a crystal ball back in January 2000.  In that light, it is quite impressive!

While in our strategy we used prior performance to switch between two exposures, momentum is more commonly applied to larger baskets of investments.  For example, a sector rotation approach may rank sectors based upon their past performance and invest in the top N.  Country or asset rotation strategies may take a similar approach.

One of the hallmarks of a momentum strategy is that it tends to be high turnover.  Momentum, as a phenomenon, tends to be short-lived.  Unlike with value, where mean reversion can take years, past relative performance tends to be useful in forecasting future relative performance over shorter horizons, like 1-3 months.  Therefore, implementation details are very important: trading costs and taxes can erode profits.

We should further point out that the crystal-ball like performance is not always the case: just a fortuitous outcome in this example.  Like the other approaches, momentum can, and will, suffer prolonged periods of underperformance. 


Carry maintains the same definition in relative performance calls as it does with absolute performance calls.  With relative calls, however, the goal is to capture the spread between high carry opportunity and low (or even negative) ones.

The most famous carry trade may be the Yen carry trade.  Prior to the 2008 recession, Japan had significantly lower interest rates than the rest of the world.  At the turn of the century, speculators could borrow $1,000,000 – in Yen – in Japan for a 0.74% rate on short-term government debt.  They could then turn the Yen into dollars and buy short-term U.S. Treasuries, which at the time were yielding 5.33%.  Such a trade would net the speculator nearly $46,000 a year.

The risk in the trade is what happens to currency exchange rates during the life of the trade.  If the Yen falls against the dollar, then the trade is even more profitable, as fewer dollars need to be exchanged to pay back the original loan.  On the other hand, if the Yen appreciates against the dollar, then our profits erode as we need more dollars to pay back the original loan.  If the Yen appreciates more than 4.6% against the dollar, the trade loses money.

We should note that the economic theory of uncovered interest rate parity predicts that in such a situation, the Yen should appreciate against the dollar so that the trade is unprofitable.  “No free lunches,” and all that.

Yet, empirically, such trades have historically worked.  Some argue that they do so because higher yielding currencies – in this example the U.S. dollar – are riskier, and hence a premium should be earned by holding it.  Others argue that the trade is self-fulfilling.  As more speculators put on the trade, they must sell Yen to buy U.S. dollars.  This puts continued pressure downward on the Yen-Dollar exchange rate, preventing the Yen from appreciating enough to erode the profits.  A third argument is that central banks have non-profit driven motives which can lead to interest rate differentials that might not exist if they were set only by borrowers and lenders.

While currency carry may be the most popular example of the carry trade, it is by no means the only way such a trade can be implemented.  For example, investors might sell lower-yielding corporate debt to buy higher-yielding corporate debt, looking to capture the spread between the two.

Carry is also a popular concept in the futures market, where commodities and more esoteric instruments – like volatility – are traded.  With a futures contract, traders speculate as to what the price of something in the future may be.  For example, let’s assume the price of gold today is $1,258 / oz and a futures contract for an ounce of gold next year is selling for $1,297 (this higher price not only captures the view of investors, but also variables like the cost to store gold, interest rates, et cetera).  As a highly simplified example, let’s assume we sell the contract today, pocket the $1,297, and note that in a year we have to make deliver on an ounce of gold.  If a year from now, when the contract comes due, gold is trading for less than $1,297, then we can buy an ounce, deliver it, and keep the difference.

This style of trade has, more recently, become a popular play for retail investors in the volatility market.  Similar to our example with gold, a futures market exists that tries to estimate the value of the CBOE Volatility Index (the “VIX”) in the future.  Estimates further in the future are often higher than those closer to the present day.  Hence, just like our gold example, selling contracts further in the future and buying them back as they get closer to maturity can be a profitable trade.

We can see how profitable by looking at the VelocityShares Daily Inverse VIX Short-Term ETN (“XIV”), which does exactly this trade.

Source: CSI.

Since XIV’s inception – and after product costs – this trade has earned over 886%.  The performance does not come free, however.  Rapid and unexpected spikes in volatility can lead to severe losses.  For example, in the two-week period from August 18th, 2015 to September 1st, 2015, the strategy lost 54%.  It is not hard to imagine how devastating an environment like 2008 might be for such a strategy.


As we mentioned before, as a style, defensive has the least concrete definition.  The general concept is that “safer” things (for a variety of definitions of safe) out-perform riskier things.  Sometimes this happens on a total return basis, but more often than not it occurs on a risk-adjusted basis.

This anomaly flies in the face over almost everything we initially outlined, where we said that extra return should be earned for bearing extra risk.  Why, then, does empirical evidence suggest that it is actually safer things that have outsized returns?  There are a variety of theories depending on the asset class in question.  Broadly speaking, there are two explanations.

The first is that investors have an explicit aversion to utilizing leverage.  Therefore, in effort to increase return, they pursue higher risk investments.  However, by collectively pursuing higher risk investments, prices of high risk assets are driven up and returns are driven down.  The second explanation is that investors collectively have a bias towards lottery-style investments, which often exhibit higher risk characteristics.  The collective pursuit of such opportunities drives up their price and down their future expected returns.  Hence, focusing on the boring and safe means a focus on investments that the rest of the market may eschew.

In other markets, the reason for the anomaly may be structural.  For example, in U.S. Treasuries, evidence suggests that 20+ year Treasuries due not earn an adequate premium above 5-year Treasuries for the extra term risk taken.  One potential cause of this is investors that do not have profit-seeking motives – like pensions – who might use longer-dated Treasuries to hedge long-term liabilities.  This behavior creates structural downward pressure on longer-dated Treasuries, creating a poor risk-adjusted return profile compared to their shorter-dated counterparts.

Ultimately, what “safe” means varies by investment.  Common measures of safety include an investment’s volatility profile, it’s sensitivity to market movements, or measures of quality (e.g. a bond’s rating).

Perhaps the most popular current example of a global tactical defensive style portfolio is “risk parity.”  In a risk parity strategy, leverage is applied to stocks, bonds, commodities, and currencies such that each asset class contributes an equal amount of risk (often measured by volatility) to the overall portfolio.  The argument made by risk parity proponents is that the approach does a better job diversifying portfolio risk than a traditionally allocated portfolio.

Since safer investments typically have less volatility, leverage must be applied.  Thus, the byproduct of a risk parity approach is that the excess risk-adjusted return empirically exhibited by safer investments to become excess total return.

Below we plot the results of one global risk parity index versus a static 50/50 stock/bond portfolio.

Source: Salient; CSI; MSCI.  Calculations by Newfound Research.  Global Static (50/50) is an equal mixture of the MSCI World Index and the Vanguard Total Bond Market Index Fund (VBMFX), rebalanced monthly.  Strategy returns assume the reinvestment of all distributions.  Strategy returns are net of all fees, including trading costs and taxes.  Returns are purely hypothetical and backtested.  Past performance is not an indication of future results.

While risk parity cannot necessarily escape the common risk factors it shares with a traditionally allocated portfolio (e.g. equity market risk or interest rate risk), by leveraging safer investments to match the profile of riskier ones, it can potentially harvest the defensive style premium.

How are these strategies implemented?

Global tactical asset allocation strategies can be implemented in a variety of different manner.

On one end of the spectrum, many of these approaches can be implemented in a purely long-only fashion with mutual funds, exchange traded funds, and individual securities.

More flexibility can be attained if shorting is allowed.  Specifically, directional bets can profit when assets lose value and relative styles can be more purely attained.

More complex approaches may begin to include derivatives – such as futures, swaps, and options – either out of necessity (e.g. managed futures), to gain leverage (e.g. with risk parity), or it may simply be the most cost-effective way to place a trade.

We will note that strategies that employ the same styles need not implement in the same manner.  For example, consider two managers that implement a country equity index rotation strategy.  The first ranks country indices based upon their valuation, and buys the cheapest 10 countries.  The second manager performs the same first step, but then uses the valuation ranks to adjust the expected returns he puts into his portfolio optimizer.  Both seek to exploit the same style phenomenon, but their portfolios may look entirely different.

Broadly speaking, when evaluating a global tactical strategy, the following questions should be considered:

  • What is the investment universe?
  • What investment styles are being incorporated?
  • How are risks being balanced within the portfolio?
  • If multiple styles are used within a portfolio, what happens when they agree or disagree with each other?
  • Are there any risk limits employed, both upper and lower?
  • Is the portfolio implemented in a cost-efficient manner (implementation vehicles, transaction costs, taxes, and total portfolio fee)?

Understand not only which styles a manager is incorporating, but how they are incorporating them, is paramount for setting expectations as to how approach will perform.

Conclusion: Appropriate Expectations for Global Tactical Asset Allocation

Value, momentum, carry, defensive, and trend styles all have a rich history of academic and empirical support.

The excess returns they have historically generated, however, have not come free.  In fact, we believe that for these strategies to work going forward, they have to be hard enough to stick with that the average investor cannot do it.  They have to be both structurally and emotionally difficult.

If they were not – if the strategies were perceived by investors as being too easy to implement and stick with – then enough money would flock towards them and the outperformance opportunity would be arbitraged away.

This means that investors who are willing to implement them should expect decades of gut-wrenching underperformance.

The good news is that diversification applies when it comes to style investing, and that combining these approaches can help smooth out the ride.  Nevertheless, investors seeking to implement these strategies and harvest their benefits should consider them as a long-term portfolio allocation, not a trade.




Corey is co-founder and Chief Investment Officer of Newfound Research, a quantitative asset manager offering a suite of separately managed accounts and mutual funds. At Newfound, Corey is responsible for portfolio management, investment research, strategy development, and communication of the firm's views to clients.

Prior to offering asset management services, Newfound licensed research from the quantitative investment models developed by Corey. At peak, this research helped steer the tactical allocation decisions for upwards of $10bn.

Corey is a frequent speaker on industry panels and contributes to ETF.com, ETF Trends, and Forbes.com’s Great Speculations blog. He was named a 2014 ETF All Star by ETF.com.

Corey holds a Master of Science in Computational Finance from Carnegie Mellon University and a Bachelor of Science in Computer Science, cum laude, from Cornell University.

You can connect with Corey on LinkedIn or Twitter.


Copyright © Newfound Research

Previous Article

Fidelity: Sale on foreign assets! Limited time!

Next Article

Market Ethos: Behaving Irrationally

Related Posts
Subscribe to AdvisorAnalyst.com notifications
Watch. Listen. Read. Raise your average.