Wednesday, August 31, 2022

Too much time in a low risk environment will lead to a bad economic outcome

The pro-cyclical performance of risky assets is well known. It is assumed that asset returns are more volatile, risky, during economic downturns; however, the behavior of risk through time is distinct from the business cycle. The business cycle and risk cycle will not always overlap. By risk cycle we mean the longer-term trend in volatility. The risk cycle may create conditions that will lead to market downturns. Hence, an investor should separate the risk and business cycle and think of them as generating two different signals. As important as the level and change in volatility is the duration of time spent in a low volatility environment. 

Research finds that perceived time spent in a low-risk regime will encourage risk-taking and lead to better economic growth, but there is a cost to growth from an extended low-risk environment. See "The impact of risk cycles on business cycles: a historical view". In an extended low-risk environment, there may be excessive risk-taking, the reach for yield which create increased financial vulnerabilities. 

Being too long in a low risk environment may cause a more severe reversal of economic growth. A Minsky moment, which will upend economic growth exists because a low-risk environment is associated with increasing complacency; the longer low-risk beliefs exist, the greater likelihood of growth reversals as investors get shocked by negative macro surprises.

The long-term trends in volatility can be measured to generate low and high-risk environments. From the risk measures across a large set of countries, the researchers created an index of the global duration for low-risk. The evidence suggest that recessions are usually preceded by a high global duration of low-risk, and a period of global extended low-risk will have a stronger effect than a local low-risk period for any given country.

A high-risk period will have an unambiguous impact on economic growth; however, the effect of a low-risk period is unclear. It will have a positive effect on growth, but as the duration of the lower risk environment extends, there will be a corresponding negative impact on growth. Too much of a low-risk environment is a bad thing.

Tuesday, August 30, 2022

Global stock and bond comovements - The stylized facts are different

How do global equity and bond returns comovements? This is an interesting question and critical for investors who are global macro traders or just global investors. Nancy Xu tackles this problem in her paper, "Global Risk Aversion and International Return Comovement". The comovement in equity markets are different from bond markets because the reaction to macro risk shocks is different.

Equity return correlations are higher than bond markets, are asymmetric, and show counter-cyclicality. When there is a negative shock, equity markets will show higher correlation. In bad times, all stock markets will move together because they all react the same to a shock that impacts risk aversion. Bond return correlations are lower than equities, symmetric, and show weak pro-cyclicality. Bonds as a safe asset will be less sensitive to macro shocks and given the differences in safety bonds will not have the same downside characteristics.  

Given these stylized facts, the author tries to provide context for why these comovements may different. For bonds, there may be a common term premium. For risky assets, there could be a global financial cycle associated with US monetary policy. Accommodating monetary policy will motivate investors to build cross-border cash flows and leverage that is reversed upon a switch in US monetary policy. Yet, these stories are not based on a common like across assets.

Instead, there is found a strong global risk aversion factor that is link to downside and upside uncertainties of global macro shocks from output growth, inflation and real short rates. Bond comovements are lower because there is a lower response to a risk aversion shock. In fact, bonds which are safe assets may not have any reaction or a negative reaction to a negative risk aversion shock so comovements will show less skew and be more symmetric. For equity markets, the reaction to a risk aversion shock will be the same and thus be stronger in down markets.

Holding a basket of global stocks or bonds will have different diversification properties because each asset class will have a different response to macro volatility shocks. 

Sunday, August 28, 2022

Macro Momentum as an extension of trend-following?

Trend-following is a price-based system, but that does not mean that trends only occur with prices or can only be exploited through prices. The underlying economic factors that drive prices can also have trends. These macro trends are what often drive price trends. Any price trend will usually be associated with trends in the underlying drivers of markets. 

If the Fed is following a policy of monetary tightening, bonds prices will usually fall. When the Fed is following a dovish policy of lowering the rates the trend in bond prices will move higher. If there is a supply shock to the oil markets, oil prices will rise. If there is stronger economic growth, the price of risky assets will increase. Similarly, an increase in risk sentiment will increase demand for risky assets. 

The macro link may not be perfect and prices trends can often be driven by short-term market dynamics, but a macro momentum model can be an effective way of trading the markets. This is the foundation for a paper from the folks at AQR, "A Half Century of Macro Momentum". Macro momentum can be another way to play market trends.

The macro momentum approached outlined in this paper is fairly simple although the actual implementation may be more difficult than a classic price-based system. Macro momentum can be broken into four different categories: the business cycle (increasing growth or increasing inflation), international trade, monetary policy, and risk sentiment. The trend in each of these macro factors have well-defined impact on markets which can be exploited. 

Increasing growth will serve equity and currency markets while it will be negative for long and short-term bonds. The impact of rising inflation will be positive for currencies although it the change in inflation forecast is at odds with classic PPP theory. Rising inflation will be negative for equities and rates. An increase in trade competitiveness will increase demand for equity indices but have a negative impact on bonds. Monetary policy tightening will be negative for all major asset classes except currencies. An improvement in risk sentiment is positive for currencies and equity indices and negative for bonds, the safe asset.

Macro momentum can be structured as either long/short through cross-sectional analysis or long directional portfolios and can be compared with trend-following portfolios. The response during drawdowns (tail events) will be different. These macro factors can proxy for what many macro managers do. 

Macro momentum is a simple way to incorporate macro themes within a portfolio based on fundamentals ideas concerning trends. Markets will often under-react to new information so following the trends in macro factors will lead to price trends that can be exploited. This is a good middle ground for incorporating the macro narrative with trend-following.

Thursday, August 25, 2022

Trend-following - If you want convexity, use a "pure" strategy

Managed futures have had a great performance run this year, but there has been significant dispersion in manager returns. A core problem is that managed futures is often confused with trend-following and all trend-following is not the same. All trend-followers are categorized as managed futures or CTAs as a regulatory matter, but all managed futures managers are not trend-followers. Managers may call themselves trend-followers but may have other strategies embedded in their funds which pollute their trend-following returns.

Cliff Asness of AQR has posted a recent research piece on trend-following to explain what it can and cannot do for investors. See "The Raison d'ĂȘtre of Managed Futures." Managers will often describe themselves as having a dual mandate of high average returns and strong returns during a market downturn, yet a pure trend-follower may have a hard time reaching this dual goal given the characteristics of trend-following. 

While not described as a divergent trading by Asness, trend-following is trading strategy that will make money when there are market dislocations or movements away from the status quo environment. If there is market stability, trend-following will not make money. Hence, returns may be high during concentrated periods that are often short-lived. There is less likely to be consistent returns because, by definition, diverges do not dominate markets. The strategy will do well over the long-run, but that does not mean that trend-following will make money all the time. There will be periods of underperformed offset by strong performance periods. The overall returns may be strong, but not during all sub-periods.

The AQR strategy is described as a pure trend strategy; consequently, it will do well when trends exist and have performance shortfalls during periods of stability. When compared to a supposed trend-following benchmark, the SG trend index, which is a bundle of many large managed futures managers described as trend-followers, the AQR strategy will have a different return pattern. 

The AQR research suggests that many trend-followers and the created index of managers, have mixed strategies that include more than just trend-following. To meet investor requests for both positive average return and downside protection, convexity, managers who often call themselves trend-followers but also include sometime like carry. 

We view carry as a convergent strategy that will do well when returns are stable, so managers may give-up some of the convexity in exchange for more stale returns through adding carry. The data suggests the SG trend has this mixed strategy. Additionally, there are portfolio structuring techniques that will impact trend-following. For example, volatility targeting, or volatility position sizing will have the impact of reducing risk exposure when there may be strong market dislocations. The trend exposure is cut at what may be the most opportune time for making money. The results are lower returns at market extremes.

Investors need to know what they are buying, and in the case of trend-following, you may not be getting a pure trend strategy because of mixed strategies and structuring. Is this false advertising? Perhaps, but a dampened trend may be what an investor wants. However, you should not be disappointed if in exchange for more stable returns you don't do as well during dislocations.

Tuesday, August 23, 2022

The failure of forecasting - Arguments invoked by skeptics

Why do investors get forecasts wrong? Why are there failures with predictions? The forecast skeptic will say predictions cannot be made because of the properties of the world and the properties of the observers. We can classify reasons for failures into ontological or psychological arguments. 

Ontological: Properties of the world 

  • Path dependencies
  • Game theory
  • Asymmetries between past and future 
  • Complexity theory 

Psychological: Properties of observers 

  • Preference for simplicity
  • Belief in a controllable world
  • Misunderstanding of probabilistic processes 
  • Aversion to ambiguity
The ontological arguments state that forecasts fail because the world itself is difficult to characterize. The past may not match the future given that the world changes. There are path dependencies that limit forecast choices. The world is complex and thus unknowable, and game theory suggests that player behavior change when faced with changing situations. 

There are also psychological reasons for why forecasts will fail. There is a preference for simplicity which in reality may not exist. There is a belief that events are controllable when there may not be control. This is especially the case when forecasting the impact of policymakers. There is an aversion to ambiguity as forecasters assume their mental models work, and there are misunderstandings for how to measure or create probabilities. Investors are generally not good at measuring and assessing probabilities. 

The skeptic will say that forecasting is hard, yet awareness of arguments for failure can help improve forecasting skill. There are areas of forecasting failure and those can be minimized through explicit adjustment of thinking. Trend-following can be effective as a prediction tool through knowing the limitations or pitfalls that may exist with other forms of predicting. Discretion can be improved by awareness of ontological and psychological properties for potential failure.  

Thursday, August 18, 2022

“We’re not overbuilt, we’re under-demanded.” - Understatement on housing market


We're not overbuilt, we're under-demanded" is one of the great understatements on the current housing market. With interest rates increasing and current 30-year mortgages at 5.5%, demand has evaporated from the marginal buyer. Simply put, many buyers think in terms of their monthly payment. If the mortgage rate goes up, then prices must come down to set a fixed monthly payment. 

The spike in rates has caused a housing demand shock.

The supply of housing has moved to the highest level in a decade. What was a housing shortage less than a year ago is now a glut. New home sales have fallen by double digits when last year supply was a problem. Existing homes were in short supply as buyers eagerly paid over the asking price. Sellers are now discounting, and buyers are walking away from signed contracts. All from a change in the interest rate environment and it will not get better anytime soon.

Global commodity and entitlement relationships in three parts

  • There is the supply of food - the measurement of what is produced for a given crop.
  • There is the command of food - the control of the distribution of commodities. 
  • There is the end ownership of food - The relationship between the food and the person who consumes it. 
  • The supply issue is a production problem.
  • The command issue is a logistics problem.
  • The end ownership issue is a distribution/pricing problem.

This framework is a variation on the economic development work of Amartya Sen on famine which can be adapted to today. There is a supply problem given current weather and production. There is also a command problem because grain from the Ukraine cannot easily come to market. There is less an end ownership problem, but it can arise if there is a shortfall of income and ability to pay from higher inflation.

When an investor wants to discuss food commodities, the conversation should walk through these three issues. You may be surprised by market action because you focused on only one price component. More simply put, looking at supply without understanding demand or logistics is a loser's game. While our discussion is about food, the same framework can be applied to energy markets. The supply exists, but the command over oil and natural gas is restrictive and the end ownership is not affordable.

The simple Einhorn bear market checklist

We have used the Steve Einhorn checklist as a simple tool for determining the likelihood of a market downturn or reversal. It is based on 5 factors: problematic inflation, hostile Fed, prospects for a recession, investor sentiment, and valuations. We are currently showing three checks for a downturn. This is down from 5 in June and early July.

Inflation is still a problem even if there has been a peak. The Fed is still hostile to markets as rate increases continue to be expected. The market may view that this will change but the wording from the Fed is clear. The prospects for a recession are still high in all market indicators but labor although there is a large difference between the household and establishment signals. Investor sentiment has turned with massive short coverings and flows moving out of cash. This is  a factor that needs to be watched carefully. Finally, valuation have come down form extremes but there are still at elevated levels. The markets are not cheap just less rich. 

The market rally is from a lessening of concerns, but that is not the same as saying there will be a market upturn.

Wednesday, August 17, 2022

What is making 2022 special for trend-followers?


What is making 2022 special for trend-followers? Many have suggested that the higher inflation environment is the answer. Others have suggested that market uncertainty which has slowed decision-making is the key, and finally there are those that say we are in a crisis from geopolitical risks. I have suggested that these are good reasons, but the answer often can be even simpler as we have discussed in our post, Turning points kill trend-following performance

If there are fewer turning points in each asset, there will be higher Sharpe ratios for a trend-following strategy. If trends last longer than expected with fewer reversals, the strategy will be a winner. The trends can be shallow, or they can be steep, but if they continue, profits will be made. Count the turning points and you will have a good idea of whether there it will be a good for trends.

From the referenced paper, "Breaking Bad Trends", there is a clear linear relationship between turning points and return. A good trend environment has four or less major turning points in a given market for a 12-month period. 

Look at some of the major markets for 2022 guidance on why this is a good year especially for long-term trend-followers. For this year, equities have had 4 major switches between 20-day and 80-day moving averages. The 10-year bond, dollar index, and corn have had one switch and oil has had three switches. This does not account for the size of the move but gives an indication of switching costs. The low switch markets have been major winners. 

The one thing that is certain for trend-following, the less trading you must do the better will be your performance. If I am working hard at trading, I am not making money. If I am bored with my activity, it is the best of times. 

Tuesday, August 16, 2022

Trend-following versus Momentum - A short comparison


While the difference between trend-following and momentum trading is clear to most active style managers, it is important to understand the distinctions and realize that these two strategies will not follow the same return behavior. Don't use these strategies terms interchangeably in conversations. Both look at past price behavior; nevertheless, one should not expect similar performance or a high correlation. 

Trend-following is an absolute return, directional strategy while momentum is relative return, market neutral strategy. The trend-following is looking to take advantage of beta direction across asset classes while the momentum trader is looking for cross-sectional opportunities within an asset class. 

Can you hold both? Yes, but do not say they are similar. 

Wednesday, August 10, 2022

Four quadrants business cycle narrative thinking


I am an advocate of the simple phrase, "you cannot know where you are going unless you know where you are". This especially applies to thinking about investing with respect to the macroeconomic environment. As a simple approach to describing the economic environment, a 2x2 matrix can be employed based on level and trend for variables that proxy for economic growth, liquidity, and risk. 

The four quadrants represent position in the business cycle: early and rebounding, mid-growth and peaks, late and moderation, and recession and contracting. The PMI can serve as a simple example. The diffusion index can either be above or below 50 which is either strong or weak, and the index may either be rising or falling. A combination of below 50 and falling is a recession and contracting regime. A low PMI but positive change would be a recovery while a high PMI and negative change would be a slowdown. This methodology can be applied to any variable tracking growth and can also be applied to liquidity and risk measures. It provides a simple narrative framework for reviewing the environment and what may be expected with returns for major asset classes.

Narrative framing with data is an effective tool for structuring investment discussions.

Measuring transitions between inflation regimes can be done

Inflation will follow regimes or cycles no different than business cycles; however, inflation cycles may not overlap or match with business cycle behavior. A good inflation paper by Kinlaw, Kritzman, Metcalfe, and Turkington, "The Determinants of Inflation" provides a useful framework for analysis of inflation regime. We will not cover their deeper inflation work using the Mahalanobis distance function which provides an attribution technique for explaining inflation regime, but we present their general framework for identifying inflation regimes. 

The authors focus on inflation shifts which is the difference between current annualized inflation and 3-year annualized inflation. Regimes are based on the form of inflation shifts. Using a hidden Markov model, the authors find that there are four inflation regimes, steady, rising stable, rising volatile, and disinflation. Inflation has been steady or rising stable for most of the post-GFC period. The 1970's were characterized by rising and volatile inflation. The current environment is now rising and volatile. This is the first time we have seen this type of regime since the early 1980's.

The features of different regimes can be viewed through macro variables that fall into different categories: cost push, demand pull, inflation expectations, monetary policy, and fiscal policy. There is a clear difference in these variables based on the regime encountered. Similarly, the returns for asset classes are very different based on the inflation regime. Investors should think about the inflation regime and the associated returns for different asset classes.  

Given these regime attributes, the authors can conduct attribution analysis from the core features on the different inflation regimes. Additionally, they are able to provide these attributes in almost real time. Their most recent version of the working paper finds that federal spending was the key driver of the rising and volatile inflation regime as of February 2022.


Tuesday, August 9, 2022

Sector rotation across the business cycle - relative and absolute performance

The asset rotation game can be played with sectors, risk premia, and asset classes such as fixed income. SSGA provides some insights on how to make these asset allocation decisions using a simple framework based on the Conference Board leading economic indicators. The LEI can be a rough signal, but an HP filter can be employed to break down the LEI into four cycle components. 

From these cycle components, performance can be measured for different sector returns on as relative and absolute basis. Focusing on recovery and contractionary phases shows significant differences in performance so that even if an investor does not get the regime timing right, there is still room for excess returns through changing allocations.  

The current environment is playing out as a contractionary regime. Better to hold Treasuries, low volatility and healthcare or consumer stables. The last month has switched the performance gains to other sectors, but a regime bias to contractionary investment may prove to be both return driven protective to investors.

Monday, August 8, 2022

Returns across the business cycle

It not a given that the economy will always rotate around the complete business cycle. It is possible that the economy will move from recession to recovery and then bounce between recovery, mid, and late cycle stages. An economy can slowdown, accelerate, and then slowdown again. A cycle is not pre-ordained nor does a specific regime have a given timeframe. This is indicative of the core problem with matching business cycle regime with asset allocation. There also is limited clarity for when an economy moves between cycles. There are no simple measures for recovery, mid, late, and recession periods. UBS provides some simple analysis on cycle uncertainty and what should do well in different stages.

The current environment is a case study in ambiguity. To some, we are in a recession, while for others we are just in the late phase of the cycle. The key problem, you cannot know where you are going until you know where you are.  


Business cycle checklist and a cycle guide

A key problem with sector rotation is determining where you are in the business cycle. A simple approach is to use some macro proxies like the PMI index, industrial production, or leading indicators as a single feature that can show level and trend. A more sophisticated ensemble approach is to employ a set of factors that can form a ranking system. 

JP Morgan has presented a three-stage model for cycle analysis based on early, mid, and late cycle. It does not measure the end cycle which is a recession. The private bank looks at nine factors and then measures which cycle stage the economy is in for each feature. The current results are from May 2022 show a mid-cycle environment given a bimodal difference between early and late factors. A closer look for a current August reading would put the environment more solidly in late cycle. These definitions are not precise, so there may be some disagreement with the cycle locations, but it provides an understandable approach to allocations. 

Mid to late cycle will place greater exposure to large cap over small cap, high quality over low quality, and a switch from value to growth. A late cycle should increase exposure to small cap, low quality, and value stocks. 

Credit cycle similar to the equity cycle - Declining fundamentals, valuations, and technicals

The credit cycle will often follow the equity cycle. Both are in bear markets. There can be a classic analysis of determining the positioning within the cycle. Robeco, in its credit outlook shows the current credit path with a weighting on fundamentals, valuation, and technicals. They argue that bear markets and market bottoms are dominated by factors other than fundamentals and valuation. We do not agree. 

The widening of spreads is driven by forward-looking fundamentals and valuation not current numbers. Fundamentals are declining based on the expected slowdown of cash flows associated with a growth slowdown and inflation. Valuations are deteriorating based on rising rates and the potential impact of QT which will create a crowding-out effects. Inflation is pushing core fixed income investors to sell duration. 

Sentiment and technicals reinforce the core problems with fundamentals and valuation. Liquidity is a growing problem because there are fewer buyers for corporate paper and dealers do not want to hold these risks.

Sunday, August 7, 2022

Inflation as a global phenomena and not just food and energy


The most recent OECD inflation data have come out, and it sends a clear message that inflation is a global problem and it is a problem beyond food and energy.  The numbers are the highest since 1990. This was before the inflation targeting by many central banks. The headline data is the highest since the 1980's. 

The shock to energy for OECD countries is greater than 40%. The shock for food is well above 10%. Monetary policy cannot solve the supply shock directly; however, central banks can slow aggregate demand. The general increase in prices is an aggregate demand problem from too much money chasing existing goods. This can be solved, but again it will impact aggregate demand.

The latest US employment numbers tell us that wages are going up well above 5% and there is a still a strong labor market. The Fed and other central banks will have to do more to quell these inflation numbers.