Trends, Countertrends, in the U.S. Dollar, Gold, Currencies

Trend is a direction that something is moving, developing, evolving, or changing. A trend is a directional drift, one way or another. When I speak of price trends, the directional drift of a price trend can be up, down, or sideways.

Trends trend to continue and are even more likely to continue than to reverse, because of inertia. Inertia is the resistance to change, including a resistance to change in direction. It’s an important physics concept to understand to understand price trends because inertia relates to momentum and velocity. A directional price trend that continues, or doesn’t change or reverse, has inertia. To understand directional price trends, we necessarily need to understand how a trend in motion is affected by external forces. For example, if a price trend is up and continues even with negative external news, in inertia or momentum is even more significant. Inertia is the amount of resistance to change in velocity. We can say that a directional price trend will continue moving at its current velocity until some force causes its speed or direction to change. A directional trend follower, then, wants keep exposure to that trend until its speed or direction does change. When a change happens, we call it a countertrend. A countertrend is a move against the prior or prevailing trend. A countertrend strategy tries to profit from a trend reversal in a directional trend that has moved to such a magnitude it comes more likely to reverse, at least briefly, than to continent. Even the best long-term trends have smaller reversals along the way, so countertrend systems try to profit from the shorter time frame oscillations.

“The one fact pertaining to all conditions is that they will change.”

                                    —Charles Dow, 1900

One significant global macro trend I noticed that did show some “change” yesterday is the U.S. Dollar. The U.S. Dollar has been in a smooth drift up for nearly a year. I use the PowerShares DB US Dollar Index Bullish (UUP). Below, I start with a weekly chart showing a few years so you can see it was non-trending up until last summer. Clearly, the U.S. Dollar has been trending strongly since.

u.s. dollar longer trend UPP

Next, we zoom in for a closer look. The the PowerShares DB US Dollar Index Bullish (UUP) was down about -2% yesterday after the Fed Decision. Notice that I included a 50 day moving average, just to smooth out the price data to help illustrate its path. One day isn’t nearly enough to change a trend, but that one day red bar is greater in magnitude and had heavy volume. On the one hand, it could be the emotional reaction to non trend following traders. On the other, we’ll see over time if that markets a real change that becomes a reversal of this fine trend. The U.S. Dollar may move right back up and resume it’s trend…

U.S. Dollar Trend 2015-03-19_08-21-35

chart source for the following charts:

I am using actual ETFs only to illustrate their trends. One unique note about  PowerShares DB US Dollar Index Bullish Fund (Symbol: UUP) is the tax implications for currency limited partnership ETFs are subject to a 60 percent/40 percent blend, regardless of how long the shares are held. They also report on a K-1 instead of a 1099.

Why does the direction of the U.S. Dollar matter? It drives other markets. Understanding how global markets interact is an edge in global tactical trading. Below is a chart of Gold. I used the SPDR Gold Trust ETF as a proxy. Gold tends to trade the opposite of the U.S. Dollar.

gold trend 2015-03-19_08-22-41

When the U.S. Dollar is trending up, it also has an inverse correlation to foreign currencies priced in dollars. Below is the CurrencyShares Euro ETF.

Euro currency trend 2015-03-19_08-23-03

Foreign currencies can have some risk. In January, the Swiss Franc gaped up sharply, but has since drifted back to where it was. Maybe that was an over-reaction? Markets aren’t so efficient. Below is a chart of the CurrencyShares Swiss Franc to illustrate its trend and countertrend moves.

swiss franc trend 2015-03-19_08-23-23

None of this is a suggestion to buy or sell any of these, just an observation about directional trends, how they interact with each other, and countertrend moves (whether short term or long term). Clearly, there are trends…

To see how tactical decisions and understand how markets interacts results in my real performance, visit : ASYMMETRY® Managed Accounts

This is When MPT and VaR Get Asset Allocation and Risk Measurement Wrong

I was talking to an investment analyst at an investment advisory firm about my ASYMMETRY® Managed Account and he asked me what the standard deviation was for the portfolio. I thought I would share with you and explain this is how the industry gets “asset allocation” and risk measurement and management wrong. You see, most people have poor results over a full market cycle that includes both rising and falling price trends, like global bull and bear markets, recessions, and expansions. Quantitative Analysis of Investor Behavior, SPIVA, Morningstar, and many academic papers have provided empirical evidence that most investors (including professionals) have poor results over the long periods. For example, they may earn gains in rising conditions but lose their gains when prices decline. I believe the reason is they get too aggressive at peaks and then sell in panic after losses get too large, rather than properly predefine and manage risk.

You may consider, then, to have good results over a long period, I necessarily have to believe and do things very different than most people.

On the “risk measurement” topic, I thought I would share with you a very important concept that is absolutely essential for truly actively controlling loss. The worst drawdown “is” the only risk metric that really matters. Risk is not the loss itself. Once we have a loss, it’s a loss. It’s beyond the realm of risk. Since risk is the possibility of a loss, then how often it has happened in the past and the magnitude of the historical loss is the mathematical expectation. Beyond that, we must assume it could be even worse some day. For example, if the S&P 500 stock index price decline was -56% from 2007 to 2009, then we should expect -56% is the loss potential (or worse). When something has happened before, it suggests it is possible again, and we may have not yet observed the worst decline in the past that we will see in the future.

The use of standard deviation is one of the very serious flaws of investors attempting to measure, direct, and control risk. The problem with standard deviation is that the equation was intentionally created to simplify data. The way it is used draws a straight line through a group of data points, which necessarily ignores how far the data really spreads out. That is, standard deviation is intended to measure how far the data spreads out, but it actually fails to absolutely highlight the true high point and low point. Instead, it’s more of an average of those points. Yet, it’s the worst-case loss that we really need to focus on. I believe in order to direct and control risk, I must focus on “how bad can it really get”. Not just “on average” how bad it can get. The risk in any investment position is at least how much it has declined in the past. And realizing it could be even worse some day. Standard deviation fails to reflect that in the way it is used.

Consider that as prices trend up for years, investors become more and more complacent. As investors become complacent, they also become less indecisive as they believe the recent past upward trend will continue, making them feel more confident. On the other hand, when investors feel unsure about the future, their fear and indecisiveness is reflected as volatility as the price churns up and down more. We are always unsure about the future, but investors feel more confident the past will continue after trends have been rising and volatility gets lower and lower. That is what a peak of a market looks like. As it turns out, that’s just when asset allocation models like Modern Portfolio Theory (MPT) and portfolio risk measures like Value at Risk (VaR) tell them to invest more in that market – right as it reaches it’s peak. They invest more, complacently, because their allocation model and risk measures tell them to. An example of a period like this was October 2007 as global stock markets had been rising since 2003. At that peak, the standard deviation was low and the historical return was at it highest point, so their expected return was high and their expected risk (improperly measured as historical volatility) was low. Volatility reverses the other way at some point

What happens next is that the market eventually peaks and then begins to decline. At the lowest point of the decline, like March 2009, the global stock markets had declined over -50%. My expertise is directional price trends and volatility, so I can tell you from empirical observation that prices drift up slowly, but crash down quickly. The below chart of the S&P 500 is a fine example of this asymmetric risk.

stock index asymmetric distribution and losses

Source: chart is drawn by Mike Shell using

At the lowest point after prices had fallen over -50%, in March 2009, the standard deviation was dramatically higher than it was in 2007 after prices had been drifting up. At the lowest point, volatility is very high and past return is very low, telling MPT and VaR to invest less in that asset.

In the 2008 – 2009 declining global markets, you may recall some advisors calling it a “6 sigma event”. That’s because the market index losses were much larger than predicted by standard deviation. For example, if an advisors growth allocation had an average return of 10% in 2007 based on its past returns looking back from the peak and a standard deviation of 12% expected volatility, they only expected the portfolio would decline -26% (3 standard deviations) within a 99.7% confidence level – but the allocation actually lost -40 or -50%. Even if that advisor properly informed his or her client the allocation could decline -26% worse case and the client provided informed consent and acceptance of that risk, their loss was likely much greater than their risk tolerance. When the reach their risk tolerance, they “tap out”. Once they tap out, when do they ever get back in? do they feel better after it falls another -20%? or after it rises 20%? There is no good answer. I want to avoid that situation.

You can see in the chart below, 3 standard deviations is supposed to capture 99.7% of all of the data if the data is a normal distribution. The trouble is, market returns are not a normal distribution. Instead, their gains and losses present an asymmetrical return distribution. Market returns experience much larger gains and losses than expected from a normal distribution – the outliers are critical. However, those outliers don’t occur very often: maybe every 4 or 5 years, so people have time to forget about the last one and become complacent.

symmetry normal distribution bell curve black


My friends, this is where traditional asset allocation like Modern Portfolio Theory (MPT) and risk measures like Value at Risk (VaR) get it wrong. And those methods are the most widely believed and used . You can probably see why most investors do poorly and only a very few do well – an anomaly.

I can tell you that I measure risk by how much I can lose and I control my risk by predefining my absolute risk at the point of entry and my exit point evolves as the positions are held. That is an absolute price point, not some equation that intentionally ignores the outlier losses.

As the stock indexes have now been overall trending up for 5 years and 9 months, the trend is aged. In fact, according to my friend Ed Easterling at Crestmont Research, at around 27 times EPS the stock index seems to be in the range of overvalued. In his latest report, he says:

“The stock market surged over the past quarter, adding to gains during 2014 that far exceed underlying economic growth. As a result, normalized P/E increased to 27.2—well above the levels justified by low inflation and interest rates. The current status is approaching “significantly overvalued.”

At the same time, we shouldn’t be surprised to eventually see rising interest rates drive down bond values at some point. It seems from this starting point that simply allocating to stocks and bonds doesn’t have an attractive expected return. I believe a different strategy is needed, especially form this point forward.

In ASYMMETRY® Global Tactical, I actively manage risk and shift between markets to find profitable directional price trends rather than just allocate to them. For more information, visit


Fact, Fiction and Momentum Investing

Fact, Fiction and Momentum Investing


It’s been over 20 years since the academic discovery of momentum investing (Jegadeesh and Titman (1993), Asness (1994)), yet much confusion and debate remains regarding its efficacy and its use as a practical investment tool. In some cases “confusion and debate” is us attempting to be polite, as it is near impossible for informed practitioners and academics to still believe some of the myths uttered about momentum — but that impossibility is often belied by real world statements. In this article, we aim to clear up much of the confusion by documenting what we know about momentum and disproving many of the often-repeated myths. We highlight ten myths about momentum and refute them, using results from widely circulated academic papers and analysis from the simplest and best publicly available data.

Read the full paper: Fact Fiction and Momentum Investing

Source: Israel, Ronen and Frazzini, Andrea and Moskowitz, Tobias J. and Asness, Clifford S., Fact, Fiction and Momentum Investing (May 9, 2014). Can be found at SSRN: Fact, Fiction and Momentum Investing