Everyone knows that bonds are rich. Right thinking people and smart beta types have looked for ways to get fixed income type results without buying bonds. At this point, it feels like bond markets are driving asset prices the world over. Negative interest rates have perversely led to bonds being used for capital gains while equity markets are being used for income. I’m pretty sure that wasn’t in the textbooks. Versions of these flows can be seen everywhere. Where bonds go, utility stocks, consumer staples, quality factors follow. Financials are the opposite of this flow, driven by net interest margins and return on equity. As bonds fall these stocks rally.
Here is a little thought experiment. Let’s compare the results of buying a basket of momentum stocks (single factor concentrated basket, price momentum) with the results of a basket tempered by volatility (2 factors, momentum (high is good), and volatility (low is good)). The difference between the two models shown below in blue, compared with the 10yr yield in red.
Source: Bloomberg; Momentum results derived from back test using Mount Lucas proprietary models
Hmmmmm… people love low volatility momentum stocks because they look like bonds. But as Minsky made clear, over time the things people buy for stability can become a source of instability. The seeds of the demise are being sown, the price moves have brought forward a lot of future income.
2nd quarter PCE is going to be 4.5% annual rate. Things are looking up. What if rates do go up. The exit door will not be wide enough.
Whatever your politics, one can’t help but be fascinated by the Trump phenomenon. One interesting aspect is how pollsters got Trump wrong. We can get insight into that mistake from Nate Silver at fivethirtyeight. To his immense credit, Silver admits he got Trump all wrong, and went back to have a look at the why and the how…. it boils down to a case of algorithmic aversion (see here and here).
Silver’s approach to forecasting, which has been extremely successful across a wide range of subjects, is based on aggregating data from multiple polls to build as big a sample as possible. It is a mechanical, quantitative, model based, statistical method. The model spits out an answer, and that is forecast. But in the Trump case, he over-rode the model. He had a bias, and in this instance he applied it to his forecast, and voila, a bad outcome. Even Silver, a devoted algorithmist if there ever was one, fell into the trap. He did not trust the model, even though it had been right so many times before.
St Louis Fed President James Bullard has released a paper detailing a revised approach to economic forecasting. It’s a very smart way of looking at the world – read it here. Briefly, he is saying that the current way of viewing the world as converging to a single state is no longer useful and instead should be thought of as a set of possible regimes the economy could visit, with the regimes being generally persistent, requiring different monetary policy responses, and switches between regimes as not being forecastable. In his submission to the FOMCs quarterly economic projections, he declines to provide a forecast for the ‘Long Run’, as it is outside his model projection range. His low projection of the Fed Funds rate over the coming years reflects his view that the present regime has a low neutral real interest rate, a switch to a higher regime is unforecastable. If it were to happen, it would cause a change to many variables – policy would not reflect a gradual shift to a single state, but would have switched regimes.
This approach to forecasting was pioneered by James Hamilton. The math is pretty complex (lots of markov processes, etc.), but here is a simple way to look at it. Suppose we have two possible states in the world, the bull state and the bear state. The variable that determines the state is unobservable, and since you can’t see it, you can’t forecast it. Suppose in the bear state that the daily returns to an asset, like the stock market, are selected from a normal return distribution with a negative mean. Conversely, in the bull state, the mean is positive. If the state variable is pointing at bear, the trend will be down, if bull, the trend will be up. The trendiness of a market is determined by how likely we are to remain in the current state. For example, if the probability of jumping from one state to another is 5%, trends are more likely to persist than if the probability were 20%. What causes the state variable to jump is unknown, as Bullard describes.
Can trend following make money in a low rate environment, and is it all bonds?
We often get asked whether trend following strategies can make money in a low interest rate environment, or in a similar vein, if trend following is just a levered long bond position that’s now run its course. In short, we think that higher rates can help some aspects of trend following strategies, but certainly should not be a driver of a long term allocation decision. The portfolio benefit of an allocation to trend following to an investor or plan with more traditional equity and credit market exposures is not solely – or even largely – driven by the fixed income exposure. Using a simple trend following model in commercial markets (commodity, fixed income and currency – we explain here why we think that’s the right approach) below we break down the sources of returns in times of crisis, and suggest an economic rationale as to why it isn’t just about bonds.
Macro trades come in two flavors, modern and classic. Modern trades are short term, liquidity driven, mean reverting market dislocations. You stare at the screen, pounce, make or lose your money, and exit. Symmetric risk, big premium for risk management and timing. Classic trades are long term, cyclical shifts in the investment landscape. Classic trades take advantage of the myopic nature markets – extrapolating the present. Classic trades have the potential to make big money, because the risks are asymmetric and the herd is against you.
We think we see a macro classic – inflation. Take a look at this study from the St. Louis Fed… https://www.stlouisfed.org/on-the-economy/2016/february/future-oil-price-consistent-inflation-expectations. Current inflation expectations imply a future crude oil price of $0 under a semi-reasonable set of assumptions. Quibble with the model if you like, but you cannot escape the fact that current market pricing anticipate little future inflation. Am I able to predict what will drive future inflation….No. Like the card counter in blackjack, however, the deck sure looks rich.
Certain funds that Mount Lucas manages may or may not, from time to time, have positions which seek to realize an exposure to future inflation. There is no guarantee that such positions, if established, will be established timely and exited profitably.
Is the Fed the world’s central bank or a domestic institution? As we see it, this is the key question for the Fed at its next meeting. The economic data since the last meeting, looked at in isolation, should lead them to continue hiking the Fed Funds rate – simply put, the unemployment rate now stands at 4.9%, and inflation has made further progress back to the target with core CPI at 2.2%. The charts below show the progress toward the dual mandate. On the employment side we look at the unemployment rate against the NAIRU measure. On the inflation side we use the sticky and flexible price series.
The MLM Index™ (see www.mtlucas.com for a description of the MLM Index™) does not include an allocation to equities. Many of our trend follower competitors do include them, and as a result we get asked all the time the reasons behind excluding them. As we have written about in previous posts (Portfolio Symmetry and Commodities are not Stocks), we believe that commercial markets like commodities, currency and interest rates are fundamentally different than equity markets, and need to be accessed in a different way, matching the economic rationale for the markets existence. Equity markets exist to fund the growth of capitalism and transfer capital from savers to businesses to be invested profitably. That’s a long only rationale in our mind, as the market participants are overwhelmingly one way. We think this is borne out by the tendency of equity earnings and prices to generally rise over time as economies grow. It also means that there isn’t a natural investment pool short the equity markets – no one has a business model that relies upon falling equity prices. One way you can see the differing utility functions is in the options market – implied volatility on puts trade at a premium to calls, as the demand for protection of a long investment book is much greater than the demand for call protection on a big real money short portfolio. Continue reading
Obviousness alert! – getting the asset class betas right is the most important first step in building a portfolio. Once those decisions are made, adding alpha is an important secondary objective. We invest along the same lines we see businesses operating in the real world – raising capital to fund growth through equity and credit markets, while managing the operational price risks that impact the running of the business as best they can. Both of these create risk premiums; markets exist to transfer them to investors. We try to balance them. To us, these two premiums are complementary, but need to be accessed in different ways. The investment risk premium funds economic activity by investing in equity and credit securities from the long side, directing capital to those who seek to expand and transfer capital risk. The price risk premium takes on exogenous input and output cost risk in commodity prices, currency movements and interest rates, facilitating hedging that allows business more certainty in operations, allowing them to focus on the core expertise. Crucially, it does this from both sides of the market, trend following long and short. Combining these is very attractive, as one side thrives on stability and generally rising growth, whilst the other thrives in times of instability. Put very simply, the investment risk premium looks for cash flows, the price risk premium looks for crash flows. We think of trend following as a beta in its own right, an important distinction between us and other more traditional long only approaches. It is a beta that most portfolios are wildly underexposed to, if they have any exposure at all. We believe in equalizing the risk between them, viewing both as having long term positive expected returns while being uncorrelated most of the time, and often negatively correlated in times of stress. When it comes to adding alpha to these balanced betas, we do all the things the academic literature leads you to – buy value, momentum persists, diversify, equally weight. Over time, we expect all of these alpha decisions to contribute strongly to the static beta approach above.
The diversification of balancing these premiums has played out a number of times over the years, and we are seeing it again as we start the new year. The stability the investment risk premium prefers is rattled by fears of global growth. This is getting offset by the price risk premium is capturing the fear as it manifests itself through flows in other asset classes – falling commodity prices, a flight to safety in bonds and flows into safe haven currencies. It has been a great example of a concept we have seen for many years, and highlights how more traditional long only approaches to investing in commodities don’t make sense. Structurally, over time, human ingenuity lowers the real price of commodities, while cyclically they move on supply and demand. They are not static long only investments, one needs to be able to take both sides of the market, taking on risk from producers and users. We think of interest rates in a somewhat similar way – a factor outside the control of a business that needs to be managed. Developed market bonds offer very low or negative real returns. Investors that leverage bonds in order to equalize volatility or returns with equity markets are running real risks – having the flexibility to take exposure on both the long side and the short side will be crucial in the coming years.
The BLS Quarterly Census of Employment and Wages, collected from unemployment claims data, provides the Labor Department with a comprehensive view of the jobs and earnings market for the U.S. economy. The first quarter 2015 survey covered over 9 million establishments employing a total of over 135 million employees—around 98% of the total population of individuals on company payrolls. In contrast the non-farm payroll series surveys a small sample, 143,000, of the population of 9 million establishments.
When I started at Commodities Corporation in 1979, we managed $100MM dollars. We had computers that sat on raised floors in climate controlled rooms. We were state of the art….and we still had 100 employees. Alas, fees were higher, but my point is that advancing data availability and data processing has been a boon to the investment industry.
But there has been an unmistakable downside to all this power and information, what my partner Roger has christened “false precision”. Everyone has lots of data and the power to manipulate it with ease. All questions can be answered with “hard numbers”. Risks can be fully calibrated (see previous post, A Note on Risk). Nowhere has this thinking reached more heroic levels of absurdity than at your Federal Reserve. From a Chairman Yellen speech, circa 2012:
Although simple rules provide a useful starting point in determining appropriate policy, they by no means deserve the “last word”–especially in current circumstances. An alternative approach, also illustrated in figure 10, is to compute an “optimal control” path for the federal funds rate using an economic model–FRB/US, in this case. Such a path is chosen to minimize the value of a specific “loss function” conditional on a baseline forecast of economic conditions. The loss function attempts to quantify the social costs resulting from deviations of inflation from the Committee’s longer-run goal and from deviations of unemployment from its longer-run normal rate. The solid green line with dots in figure 10 shows the “optimal control” path for the federal funds rate, again conditioned on the illustrative baseline outlook. This policy involves keeping the federal funds rate close to zero until late 2015, four quarters longer than the balanced-approach rule prescription and several years longer than the Taylor rule. Importantly, optimal control calls for a later lift-off date even though this benchmark–unlike the simple policy rules–implicitly takes full account of the additional stimulus to real activity and inflation being provided over time by the Federal Reserve’s other policy tool, the past and projected changes to the size and maturity of its securities holdings.
Lord have mercy! Can she really think this can work? Is there any Fed model that caught 2008, 1998, 1987, or any other market moving event? This is the event horizon of the false precision black hole. But don’t worry, just throw a bunch of junk mortgages in a bucket, stir, and voila – AAA. It’s all right here in this spreadsheet.