Machine learning - what it is and is not for investors

I have been seeing more managers use machine learning tools to help with the investment process. This is an important advancement and will be very useful in helping generate better return generating engines. However, there are may investors who do not known what machine learning is or what it should be able to do. Simply put, machine learning is having an algorithm learn without explicit programming, a process of improvement with experience from new data. It is the training of a model for data that can be generalized for decisions against some performance measure. 

I have listed some key ideas of what machine learning is and is not for an investor who is just being exposed to the concept.

Machine learning is not artificial intelligence.

Many think that machine learning is a sub-discipline within AI; however, there has been a large divergence between machine learning which is grounded in data and statistics and AI which is focused on logical systems. AI, for many, never realized its potential in the investment area. We do not believe that the same failure will exist in the machine learning field given the different goals and objectives.

Machine learning is not data mining.

Data mining focused on finding relationships in large data sets that were not able to be easily discovered or are not readily apparent. Machine learning potentially uses large data sets to train systems to make predictions. The elements of machine learning include: data mining, statistical inference, and prediction.

Machine learning is not a black box.

Many AI models have relied on neural networks which can seem like a black box which cannot be understood by the manager or investor. There is talk of a “hidden” layer which adds mystery. Machine learning can be closely directed or supervised in order to apply what has been learned to new data, or it can be unsupervised in order to draw inferences from data.  

Machine learning is not new.

Machine learning was first developed decades ago with many of the statistical techniques used coming from older approaches toward inference and prediction. The more recent development is that cheap computing and large data sets has allowed for more and quicker learning.

Machine learning is not testing endless alternatives.

Machine learning does not mean that a computer and manager is going on an endless hunting expedition for an over-optimized result. In fact, the value of machine learning is the ability to blend and weigh many alternatives which could not be done in the past. 

Machine learning requires statistical/programming/market skill.

Fundamentally, the manager who uses machine learning combines strong statistical foundation with programming skills to look at a  wide set of alternative which can be weighted or excluded. Nevertheless, there still needs to be a sense of market to help learning and understanding.

Machine learning will find relationships you did not expect.

Machine learning is not just data mining which is explicitly looking for relationships within a large data set; nevertheless, machine learning may be able to weigh different model alternatives and find combinations which would not be thought of through the normal process of simple hypothesis testing. There is learning without explicit programming.

Machine learning does require a lot of computing power.

The machine learning by its very nature will look at many different combinations of data and learn how to update models when new information is introduced and patterns found. The process of stepping forward and determining the impact of new information is computing intensive.

Machine learning can be either supervised or unsupervised.

While machine learning can be unsupervised and free form when looking at model alternatives, it can also supervised whereby the learning is very specific and directed to apply what has been learned from new data.

Machine learning will make you a smarter investor.

When machines learn, the modeler learns. One of the key advancements of using machine learning is the ability to find dynamic relationship in data which are not immediately obvious and make predictions.

Structural checklist for investment headwinds

Systematic models are very useful and important for disciplined investing, but the

percentage of the variation explained by most models in asset markets is relatively low especially over short horizons. The question for most managers, even quants, is determining how to deal with this percentage that is unexplained. There is no simple solution but applying a structural headwinds/tailwinds checklist may provide a good first pass for addressing the problem.

A structural headwinds or tailwinds checklist groups or categorizes issues that may provide a tilt to returns. These tilts on expected returns may lead investors to make a tilt or base adjustment to asset class allocations. Instead of starting with a base allocation of zero, there could be a negative or positive base allocation.

There could be more categories to this checklist, but these eight may get any discussion started. Some of these issues can be quantified, but we believe they often cannot explain short-term variation except if there is a shock. These factors can have an impact on returns through impacting the risk premiums in markets but only over long horizons.

Structural headwinds - A checklist

• Demographics - 
• The era of aging is upon us and it has an impact on capital flows and savings rates; just ask Japan or Europe. Demographics may include issues like the flow between rural and urban areas in China. It will drive return patterns even though it will not affect short-term volatility.
• Government -
• The type of government that is in place will affect investment options. Government impacts could include gridlock.  Venezuela is a perfect example for where government matters. The same can be said for Argentina or Russia.
• Regulation - 
• The Dodd-Frank regulation has an impact on returns but it is hard to model the impact directly. All of these rules on banks impact credit and the financial sectors. More regulation will have a greater impact on small cap stocks.
• Elections -
• BREXIT will impact all of the globe. The US presidential election may have a profound change on trade and global relations. The election dates are well known and may outweigh other model factors.
• Geopolitical risks (war - terrorism)-
• These risks usually lead to shock effects, but the changing probabilities of geopolitical risks will impact all returns; nevertheless, it is hard to include in any model. 
• Global trade -
• This factor refers to the overall integration of capital, labor, and good around the globe and not the trade balance of any one country.Globalization will impact capital market integration will effect the correlation across markets. 
• Climate - weather-
• While there has been much talk about climate change, the impact on investment returns is less clear-cut. Obviously, there are weather events that impact returns, but these can be diversified. 
• Technology - 
• Technology can provide a boost or a drag on specific industries. In the global macro arena, the impact of technology is less clear but the impact of technology on finance is real and does affect liquidity which is being priced in the markets.

Even if these factors are not explicitly modeled, they will affect allocation decisions and should at the least be catalogued and discussed.

Managed futures - something going on with performance

It is always worth monitoring the long view with trends. Beyond the last quarter or year, there are return patterns that show longer-term changes in style performance. Take a look at managed futures. There was a long period for which it was out of style, but that has changed markedly. (Saying a strategy is out of style is the nice way of saying it has performed poorly.  In style, means it is doing well and investors are chasing performance.) 

We fitted a long-term linear trend through the performance curve of the SocGen managed futures index. The times of strong or poor performance from trend are obvious by looking at periods above or below the trend line. A fitted polynomial through the data can find periods when the slope of returns was rising or falling and provide a smoother view of performance.

There are clearly long periods of strong and poor performance. There is an uptrend in performance through the Financial Crisis, but the slope turned down during the  post-crisis or QE period. Nevertheless, the second half of 2014 marked a new period of performance. This new era is completely different than past periods. (We looked at log value and see the same pattern.)

Of course, we fitted the entire period which would not be available to investors, but the story holds. If an investor looked at holding a managed futures basket when the non-linear curve slope is rising for the index and selling during periods of declining slope, you would have have positive style tailwinds. This is not a model,  but may be a good rule of thumb. 

Stop the mistakes - play the odds from a process

When you repeat a mistake, it is not a mistake: it is a decision.
- Paulo Coelho

So what is an investment mistake? If the odds are 60/40 in your favor and you lose, is that a mistake or are you just unlucky? In this case, a bad outcome is just bad luck. If you calculate the odds to be 60/40 against you and you have a gain is that a lack of skill or just being lucky. These are important issues to consider if you want to measure or show skill. 

A mistake is not being rational or being inconsistent with an articulated decision process. If there is a process and the decision-maker deviates from the process, it is a mistake regardless of your luck. If you don't have a well articulated process, you will not be able to say that you made a mistake. A more fluid process allows for ambiguity and thus makes it hard to admit a mistake even during bad performance periods.

Given there is a process, the real issue is whether the decision-making model is wrong. If you are rationally following the model, the focus is on whether the model of market behavior is incorrect. If you consistently follow a disciplined process but if the model's predictive power is flawed, there is a mistake. Unfortunately, most models of asset prices can only explain a small portion of the variation in prices. Hence, mistakes can be repeated and turn into bad decisions.

Being systematic is just a start. It is the predictive process for which discipline is applied that really matters. Even here a trader faces a severe problem because mistakes are generally more frequent than success. A higher success ratio does not mean high returns. Success could be less than 50% yet produce higher returns. You can make repeated mistakes and still profit through controlling risk. A flawed model can still be effective if the trader cuts his losses. 

Mistake happen but decisions matter.

News, trading, and decision-making - Type I and Type II errors

Investment decision-making is a combination of assessment of information and action. Analysis and trading. If you are right, there is profit but if you commit an error, there is a loss. A good analyst who cannot act with a specific recommendation is not worth much. A trader who is not a good analyst is just a busy investment professional without purpose. From a macro perspective, how traders act and the type of errors they are willing to make will have an impact on price behavior.

When a trader receives some new "news", he has a set of choices which will either lead to profit or an error. Yet, all errors are not the same. Traders can make type I or type II errors which will have an impact on the type of trading conducted and affect the market in aggregate. Traders have to assess the probability of the type of error to be committed no different than any statistical test. 

Walking through a trader's decision tree will help explain the problem. New information may enter the market. For example, it could be a macro announcement on the unemployment rate. The trader then has to make a determination to either act or not to act based on this new information. Action can either lead to a gain or a loss. Similarly, no action can either be the right trade or not. 

If the trader is correct, there is no error and there is a positive gain. However, if the trader is not correct, he commits a type I error which is a false positive. He thinks prices will react from this news and decides to trade. He expects an action that does not occur. He will then have to reverse the trade given this type I error.   The trader detects and effect that is not present.

On the other hand, the trader may conclude that no action should be taken from this information. If there is no price reaction, this was the correct action. If he is wrong and there is a large move, the trader committed a type II error, a false negative. In this case, the trader does not have to reverse his mistake, but it may require a catch-up trade. There is a failure to detect an effect that is present.

A trader will get more false positives from acting too rashly while there will be more Type II errors from passivity.  More type I errors will require more reversals of trades. In aggregate, if the market makes more type I errors, there will be more mean-reversion. If the market makes more type II errors, there will be more trends or momentum.  Type I errors are costly because transaction costs are taken and capital is engaged. More Type II errors from lack of action are an opportunity cost. Good traders have to think about the balance between these errors. 

Performance and story-telling - A good due diligence combination

One of the most perplexing and important issues for any hedge fund investor is manager due diligence. Investors have gotten very good at operational due diligence in the post-Madoff era, but the same cannot always be said about skill assessment or investment process due diligence. 

There is no doubt the quality of skill assessment has improved over the last few years. The ascent of factor-based analysis has done a good job of better describing the risks taken by managers. Skill assessment has also done a better job of separating alpha from beta for many hedge fund styles. Still, more work is needed in this area. There is still too much reliance on performance screens versus an assessment of a manager's ability that forward-looking.

We think a simple two approach can be helpful. While many investors implicitly do this analysis, it is not structured within the due diligence analysis.  The two step approach is to to look at performance and the story that is told around performance. We believe the power of story-telling or describing the investment management process has the power to explain. This power of explanation is critical in being able to differentiate skill from luck. 

We will provide more detail on what is required for good investment story-telling but our simple matrix can describe our two-step approach. Good performance with a good story will show skill. Positive performance with a poor story concerning the investment process signifies luck. Similarly, negative performance with a good story would suggest an unlucky manager. A negative story with poor performance may suggest limited talent. 

Good story-telling may be viewed as just good marketing; however, it can be much deeper if the due diligence is conducted properly. The critical issue for due diligence is determining whether screened performance for a specific period of say three years is a true indicator of future performance. There is little doubt that skilled managers will perform well, but it is unclear over what horizon or whether recent behavior will carry-over in a different environment. This is why the qualitative story is critical. Story-telling about how investments are made, risk is assessed, and models built is a differentiator for picking managers.

Value and momentum - why do they both work?

Two of the most important factors for investors are value and momentum. These are factors that can both easily be exploited and accessed in the market. Value indices are relatively easy to create using a simple set of rules. Momentum is also easy to create through rules or through specific fund styles. Nevertheless, the stories used to explain the excess returns in value and momentum are very different.

A simple matrix based on five criteria can be used to explain the difference between these two factors. The approach of the  value factor is to look for cheapness or richness relative to other securities in a portfolio. For momentum, it is looking for assets where there have been high gains versus declines. 

For a risk-based story for the excess returns associated with value, there is the view that investors are compensated for risk from firms that may be out of favor. In the case of momentum, there is the story that excess returns are associated with economic trends, the business cycle, or behavior within an industry. Investors are also compensated for the risk of a sharp reversal.  The behavioral story for value is based on the idea that the assets are mispriced. Prior risks cause investors to miss the true valuation. The momentum story is that there is a slow adjustment from the inattention of investors. Markets do not react fast enough which may lead to overshooting and then reversals.

With respect to efficiency, value buying attempts to move markets to fair value and thus make markets more efficient. Momentum is likely to be driven by fundamentals that are moving pricing away from existing levels. Momentum in some cases can make markets less efficient by reinforcing behavior which leads to overshooting.

Investors exploit the value factor through their styles of managing their portfolio which we call convergent trading. There is movement or convergence back to fair value. Momentum is inherently a divergent style because there is a movement away from current equilibrium. 

The explanation for returns for the value and momentum have been classified as either risk-based or behavioral based. These explanations are fundamentally different. Hence, a value factor based portfolio will be uncorrelated with a momentum factor based portfolio.

These differences are important when constructing a factor diversified portfolio. For example, combining managed futures with a equity long/short will create a very different and unique profile given the different combination of factors.

Factor-based investing in fixed income

Fixed income investing can be broken down into factor-based decisions. The risks of fixed income can be described through three main factors, the Treasury rate, growth, and volatility. The Treasury rate explains the variation with the bell-weather 10-year, the growth factor looks at a combination of the global equity and high yield excess returns, and a volatility factors as measured by the change in the VIX index. This is a simple approach but gets at the heart of the factor approach. Holding a credit-based portfolio is very different than a the Barclay's Aggregate index. With credit or emerging market fixed income index, you are holding a growth focus while the Agg is just rate focused. Volatility exposure is not important except for credit and emerging markets. 

This work is outlined in  "Factor Approach to Fixed Income Allocation" in the Journal of Investing, spring 2016. The factors do not sum to one. When tested as a three factor approach which are forced to sum to one, the volatility factor drops to zero. 

All fixed income sectors are not alike and a balance between rate and growth will require a balance between credit and a more general bond index. More can be done than what is presented in this very simple factor approach but it serves as a good first pass on the issues associated with a fixed income factor approach.

Multi-asset class investing - more important than security selection

There is a growing trend that investors should focus more on multi-asset class investing or the asset allocation decision over security selection. This is an important thinking for most investors. Don't worry about the elusive alpha. Focus on diversification and your betas. This is where the bulk of your returns will come from and where investors face the most risk. If you get the asset allocation wrong, investors will need a lot of alpha to make up for the difference.

One can view risk parity as one innovative approach to this allocation issue and factor-based investing as another solution to the problem. Risky parity says diversify across multiple asset classes but make the allocation through contribution to risk. Factor-based allocation looks at the underlying driver of the asset class. The general scheme is to move away from security selection and focus on sector selection. This can be done either on an active or passive basis. 

Certainly there is a mismatch of resources in asset management concerning this issue. A recent writer, Pranay Gupta, noted that 80% of the resources goes to security selection even though it may represent only 10-20% of the returns for most portfolios. The research shows that strategic asset allocation is more important even though it does not get the resource focus. The driver of portfolio performance in the first quarter was based on the ability of investors to hold more stocks after the announcement for more liquid monetary policies around the world than any security selection.

The difficultly of the problem does not change the result that asset allocation matters. Some may view that it is harder to predict or manage the beta risks for asset classes, but the movement to factor-based risk premiums allows for multi-class asset allocation that is diversified and manageable.  In a broad sense, global macro hedge fund strategies attempt to dynamically address the issue of asset allocation. 

The multi-asset question is more consistent with how investors think about making investment choices. Most investors do not think about the excess return of stocks versus the risk free rate. The discussion is focused on whether equities will do better than bonds, or whether credit will do better than Treasuries, or emerging markets will do better than developed market stocks. The context of most investment choices beyond the security selection is relative to another asset class. Hence the focus on asset classes is imperative. 

Driving FX markets - common risk factors?

What drives the FX markets? This is a critical but complex issue for any investor who needs to look at this asset class. The usually approach is to develop a model with a some key fundamental factors like rates, inflation, or money and then see what the reduced form empirical relationships show significant relationships. However, this problem can be looked at with a more primal lens through the use of principal components. Using this approach, it is possible to determine how many common factors may drive currencies.

Researchers have found that there are two principal components that can drive these markets. See "Common risk factors in currency markets"  by Lustig, Rousssanov and Verdelhan. The main first principal component can explain about 70% of the variation is related to risks versus the dollar. These are the usual risks investors often thing about with respect to currencies. What is more interesting the second principal component that seems to do a very good job of explaining the cross-sectional variation in currencies and explains about 12.5% of the variation.

This common factor  explaining excess return is sloped with respect to the forward discount or interest differential. Simply put, excess returns are associated with high foreign interest rates. This is the classic carry trade factor.  What is important with this factor is that it seems to be related to world risk and is counter-cyclical. It is like the risk premium found in stocks and bonds. When there is higher forward discounts, there will be higher excess returns from the high foreign rates, but this is associated with pricing of world risk. 

This story makes sense that carry will do poorly if there is a recession or if there is an increase in global volatility. You are not getting a truly unique return pattern but one that will be sensitive to the business cycle. Excess returns from currencies with high interest rates need to compensate for the risks associated with "bad times". Low rate currencies will be safe havens. This work does to mean avoid carry trades or currencies but it better explains that you are receiving excess returns for specific risks. 

Hedge fund performance mid-month - mixed

There have been more news stories this month about the poor performance of hedge funds in the first quarter of 2016 and the fact that there are rumblings that investors will pull money from hedge funds. Hedge funds have to get their act together on performance, fees, and the value proposition. 

The first half of the month has seen some bright spots as measured by the HFR indices.  Directional hedge funds and those that generally  have more beta exposure did better. The market directional index posted the best performance doing even better than a stock index. The only style that showed a drag on performance was the equity market neutral which was down 114 bps. Almost all of the hedge fund styles beat the long bond. 

Now, it is not always a good idea to compare hedge fund returns over a short horizon against equity and bond benchmarks, but unfortunately, many decisions are based on relative performance. Except in very strong up markets, the expectation from many investors is that the cumulative alpha will allow for better risk adjusted performance versus a beta index. If there is no strong alpha value, investors think they are just paying high fees for lower beta. This is the investor mentality that has to be overcome by hedge funds. 

BREXIT is the trade of the year

The BREXIT issue may be the trade of the year. What does that mean and why give it this view? 

1. This is a big event - Not only will this have huge impact on Great Britain and its place in the European and world economy, there will have a spill-over effect to the rest of the EU. This is the type of event that will shift rare event expectations and increase tail risks.

2. There is a specific date when uncertainty will be resolved. There is not an issue of when it will occur. We know the vote date.

3. The probability of the vote can be tracked and measured. Handicappers now say there is a 24% chance of an exit. The poll numbers look like it is very close. The risks of a bad outcome after June are still not clear, but the vote can be watched and given precision.

4. The market has already spoken that this is a big deal and talk is generating more interest. The IMF just stated that there will be severe global damage if the Brits leave. 

5. Financial institutions will have to square their risk books so there will be a liquidity effect and there will be a risk premium for those that hold these risky assets. This has started and will continue as we get closer to June. There will be a flight to quality effect regardless of the vote outcome. 

There may be better trades that develop through the year and ex post we find had a large pay-off, but this is the most measurable big event trade of the year.

 

FX as an asset class? - what matters are factors

I visited the FXWeek North American conference across the street from my office. It is interesting that the issue of FX as an asset class is still being discussed as a topic. This issue just will not go away. While I think of it as an asset class, there are still many who are skeptical; nevertheless, we can change the focus from asset class to factors and then foreign exchange becomes more unique and useful. If the idea is that you would like to buy unique investment factors, currencies are good place to gain return and diversification.  In fact, some of the well-known factors that are applicable in equities or fixed income also apply to currencies but with a different return profile uncorrelated with other asset classes.

What are the factors that can be the focus of foreign exchange? The usual suspects apply to foreign exchange - carry, momentum, value, and volatility. Exploiting carry in FX is different than in the fixed income markets. It may be related to global risk but it has a different profile. It is a different way of playing carry. Momentum in currencies has been a long-term winner for investors which is different than the momentum in equities or commodities. FX value is much harder to measure than in other asset classes but it can be used as a factor. The same can be said about variance. All of these factors have been well-established in currencies, so if you want factor diversification you get it in currencies.

Providing uncorrelated returns on similar factors can be good way of determining the uniqueness of a an asset class. One can change the argument about asset classes and focus on different ways of expressing factor bets.

News may lead to volatility and disaster uncertainty - a priced variance premium

News headlines will impact market returns and volatility. Headlines often represents uncertainty and can increase the fear or probability of rare disasters.  More news headlines that generate investor uncertainty will lead to higher price volatility. 

Headlines will change return expectations and cause investors to switch their views on the chance of markets being in a good or bad state. This will lead to greater deviations in prices. An increase in volatility will mean that return premium have to increase to compensate investors for the added risk from uncertainty. The short-term impact between risk and return can be complex, but longer-term links between news implied volatility and return is actually well-defined. 

This link is especially the case when the news is focused on low probability potential disasters. There has been strong research work that when the probability of a rare disaster or left-hand tail event increases, returns will have to also increase to compensate for risk. Rare disaster risk has been used as an explanation for the excess return premium in equities.

Asaf Manela and Alan Moreira in their paper, "News Implied Volatility and Disaster Concerns" focus on key headline risk using machine learning to tease out text-based measures of uncertainty in newspaper headlines. Their focus is not on stock volatility but changes the chance of a rare disaster. 

Their research work finds a positive link between their news implied volatility index and return patterns. There is a variance premium with respect to their news measures. Of course, this does not mean that you should just ride the wave of negative news with a passive portfolio. Investors will be harmed by disaster shocks or changes in disaster expectations, but returns will then rise in the future to offset this risk. The behavior of the markets in response to war, government redistribution changes, or financial crises is well-defined and linear. Higher news uncertainty will mean future higher returns. 

Fading volatility is a strategy that works

Volatility matters. Portfolio efficiency focuses on the return to risk trade-off. There has been a focus on low volatility strategies, but just as important is the idea that when volatility is high or increasing, investors should be walking away from risk. Adapting or adjusting to volatility is a dynamic concept. In simple terms, investors are often not compensated for shocks or increases in volatility. The low volatility argument is that investors are not compensated when volatility is high. Investors do not receive enough return compensation for either level of changes in volatility to keep their Sharpe ratios stable.

Given this viewpoint, along comes a very interesting paper, Volatility Managed Portfolios, by Alan Moreira ad Tyler Muir. The authors show that dynamic adjustment of portfolio exposures with respect to volatility will add to the portfolio's overall Sharpe ratio. If volatility is high, cut the risk exposure to that asset or factor. 

This is contrary to traditional thinking which says that there is risk return trade-off such that higher risk will be offset by higher return. What the authors find is that the response of return is slow relative to a shock to volatility. If volatility spikes, cut the exposure. As the volatility comes down, you can go back and increase exposure and you will still benefit. This applies even after accounting for well-known factors. 

What is very interesting is that this works even during periods of recession. Volatility will increase during periods of recession or "bad times". You will also be compensated for risk by buying risky assets during bad times but the timing of these two relationships is not the same. Hence, volatility timing can add value and you can still hold the relationship that returns will be higher in periods of recession.

This makes sense although you need to see the numbers and get comfortable with the fact that changes in risk do not immediately translate into changes in return. Volatility is forecastable over relatively short horizons while returns are not. This means that investors can exploit volatility shocks. 

On a risk-adjusted basis, investors are worried about risk and return. If you can make good volatility choices and cut high volatility, you are able to increase the mean variance ratio.  This is research that is easy to implement and have a meaningful positive impact for investors.

Volatility on downtrend - volatility spikes tied to liquidity

Asset price volatility follows a pattern of spiking only to then gradually decay. This is the essence of much of the GARCH modeling revolution. The patterns for volatility behavior is well-defined. The key issue is what causes the spikes in volatility. A simple answer is market uncertainty, but it is hard to determine what is uncertain versus what could just be a surprise. Surprises are unanticipated but may not be the same as uncertainty. Surprises will cause spikes in return, but they may not always lead to persistence in volatility. 

Many volatility spikes are related to changes in liquidity as defined by money in the credit channel. A decline in money through increases in rates will cause reduction is risk-taking and change the discount factor used to price assets. This will lead to higher volatility which will decline as the market adapts to the change in liquidity conditions. Increases in liquidity will not generally lead to increases in volatility. Decreases in liquidity will lead to volatility spikes because risky assets will be sold. Increases in liquidity will dampen volatility. 

Stock volatility has declined after the surge from the Fed rate increase. Since January there has been a dampening in volatility associated with increases in liquidity by other central banks and the potential delay by the Fed. Now the null hypothesis on volatility spikes is that there will always be a dampening after a volatility surge, but the increases are greater on liquidity decreasing shocks. The speed of the decline will be associated with the reversal in liquidity tightening.

With this view of the world, volatility will stay low until the next liquidity spike. 

Asset management as cooking, asset classes as food, and factors as nutrients

Just like ‘eating right’ requires you to look through food labels to understand the nutrient content, ‘investing right’ means looking through asset class labels for the underlying factor risks. It's the nutrients in the food that matter. And similarly, the factors matter, not the asset labels. - Andrew Ang, Columbia University 

The Ang quote has been making the rounds of financial blogs and news for some time. I like the analogy. Asset classes are like different foods. The real value of the food is in the nutrients. We eat the food to ultimately get at the nutrients. 

Unfortunately, we often do not eat the nutrient directly. The foods and nutrients are balanced in the meals we eat. We like some foods. We have aversion to others. There are good meals prepared by skilled chefs and there are bad meals where we do not get any value. The difference is often with how the food is prepared. Asset management could be considered the cooking or the preparation of the meal. 

Asset management is about balancing risk factors and finding or generating alpha. The combination or choice of foods and factors and their preparation create alpha. The nutrients are out there, but some managers are able to bundle or find them better than others. Similarly, a bad chef can have access to foods and nutrients but actually deliver them poorly. We are underwhelmed with the preparation of some of our meals.

There is a price for food and those nutrients. We will not pay a premium for poorly prepared food and we should not pay a premium for poor asset management. Those managers that can deliver the right balance of factors should be paid a premium.

You never want to go too far with any analogy, but the skilled manager who effectively uses asset classes and factors is no different than a chef who uses well food and nutrients. 

Asset class diversification is good, factor-based may be better, but strategy diversification best

After attending a conference that highlighted some of the latest research on smart-beta and factor-based investing, I can say that this new focus is taking the industry by storm. There was no discussion about allocation to asset classes except in the context of benchmarks and the "old" approach. There is no questions that factor-based approach is a more nuanced view on risk allocation, but there is going to have to be more education on how this research can be implemented and what are the right factors. 

One of the clear advantages of factor-based investing is that it can better target the unique (orthogonal) risks in the portfolio. Unfortunately, there is growing number of factors that are being discussed as potential risks. The market has moved well beyond a Fama-French three factor model approach, but how many risk factors are out there and how many are truly needed is very much up for debate. there is no set standard don what is a good factor and I believe the factor used will differ based on the type of manager you are. A global macro manager will be looking at macro risk factors cross equities while a long/short equity manager will have a focus on cross-sectional differences within a set of country-specific stocks.

Regardless of the discussion on what factors, the focus on diversification across risk factors is better than just asset class diversification. Too often when there is a common shock to the markets correlations across asset classes move to one. This movement to a correlation of one is the disappointment with traditional diversification and a key driver for factor-based work especially with respect to macro factors.

Factor-based work also better identifies and isolates the inherent risks in a portfolio whether size, profitability, value, momentum, growth or inflation. Not until risks are identified can they actually be managed.

However, it is the dynamic use of factors where there is the potential for true excess returns. Factors may be time varying, so the management of factors is the next level of advancement. The management of factors could be called strategy. Hence, we believe that strategy is the bets form of diversification and allows for the greatest amount of diversification.

Andrew Ang of Columbia U. has used the analogy that asset classes are meals or food while factors are the nutrients in the meal. If that is the case, then strategy is the cooking of the food and nutrients. A great chef can make your meal better.

March hedge fund performance moving higher, but disappointing first quarter

With equities moving higher and credit spreads tightening, hedge funds with equity beta exposure  showed strong returns. The exception was with the CTA and global macro indices which posted the largest declines. The best hedge fund strategies included market directional, fundamental growth, distressed, and emerging markets. It was a clear risk-on environment for the month.

Nevertheless, most hedge fund strategies did not do well during the first quarter of the year. The only strong performer was the systematic CTA index which posted returns over 6%. The directional equity hedge funds have return differentials of over 12 percent relative to the CTA leader.

The first quarter just provides more evidence that all hedge fund strategies are not alike. If there is more financial turmoil, the divergent strategies like managed futures will perform well. This strategies take advantage of market turmoil. Convergent strategies which focus on markets remaining well-behaved or those that have higher exposure to market beta will be harmed. 

What was a true disappointment is that many hedge fund managers were not able to protect wealth on the way down and then gain returns when the market reversed. Most strategies were slow-footed at adjusting market exposures. It will take some work to match CTA's if the divergent strategies hold their gains. 

Global macro themes on one page

The major theme for the second quarter will once again be focused on the Fed and rate increases. A market-dependent  Fed creates a feedback loop between market prices and Fed action. It seems the Fed is closely watching market behavior and will err on the side of caution if there is any sell-off or price disruption in markets. The market-dependent Fed has supported emerging markets and a risk-on equity environment. The markets have to read the thoughts of this dovish Fed and determine when the next rate increase based on any market feedback.  

Managed futures sector review - liquidity gaps and trend reversals

What was the main cause for the negative performance with managed futures for March? The reversal in sentiment across bonds, rates and currencies is the easy answer. Bonds were selling-off in early March as a result of the strong equity rally. The switch from less risk to riskier assets dominated the movement in financial futures. However, the change in sentiment mid-month from the belief in a more dovish Fed caused bonds and rates to rally and reverse the bond outflows. After the bond rally from the earlier equity sell-off, traders shorted positions in March only to find themselves wrong-footed toward central bank sentiment. Long-only bond returns were flat but active traders were hurt by the intra-month gyrations.

The dovish Fed sentiment carried over to currencies where trades based on the high rate differential in favor of the US were hurt. Currencies are expectational markets and the forward view is that central bank policies will converge not diverge. The dollar rally continues to be on hold. 

The oil market continued to rally but the end of the month showed some downward price pressure based on the view that further supply cuts will not be forthcoming. Nevertheless, longer-term capital projects have been cut. This will provide a price floor.

Precious metals have been a big winner for the quarter based on higher inflation, negative rates and continued market uncertainty, but momentum slowed by the end of the month. Base metals which strengthened earlier in the quarter have become at best more range-bound based on lower but stable China growth and global growth that will be positive but modest. Commodities have moved on market-specific economics and not any common factor. The value of commodities outside of oil is their unique price behavior relative to traditional assets. 

The only strong trend was in US equities. We have given sideways directions to other sectors because of the strong reversals during the month or the slowdown in momentum. A continuation of second half trends from March could mean better performance for CTA's, but positive bond and equity correlation is not usually expected for prolonged periods. 

Longer-term versus short-term managed futures performance

A key driver for holding managed futures is crisis beta, the low or negative correlation during times "bad times" versus traditional assets. The crisis beta is a direct result of the divergent strategy of long/short trend-following. If there is a market dislocation, the trend-following may short the assets that are doing poorly and increase long exposure to those assets that are doing well. This will dominate long-only investing that is not fully diversified. 

Unfortunately, there are few guarantees of when trendS will occur and there is the potential for momentum crashes when trends reverse. Simply put, trend following will do well during bad times but there is a cost during normal times or period of trend reversals. You have to pay price for the benefit of trend-following. 

One way of reducing the risk from trend-following is to diversify the strategies employed within managed futures. Instead of only using trend following, diversification can be achieved through investing in short-term strategies. These strategies will provide diversification versus traditional assets but will also be less correlated with longer-term trend managers. Short-term strategies could still be focused on trends but the shorter time-frame makes it less correlated to the strategies that are looking for longer-term trends. 

Time horizon matters as a diversification tool.  

March is a perfect example of the difference in time horizon. The SG managed futures index declined by over 3 percent while the short-term traders index declined just over 1 percent. The short term index performed better in the first quarter albeit it did worse in 2015 and lagged when the large trends in the first two months appeared.

 A comparison between the two strategies since the beginning of 2015 shows that both happened to end up at the same spot, but the short-term index did it with less volatility. This is the power of looking to exploit shorter-term trends or counter-trends.  A mix of time diversification can go a long way to smoothing returns.