Capital Market Expectations, Part I: Framework and Macro Considerations

This is the first of two readings on how investment professionals should address the setting of capital market expectations. The reading began with a general framework for developing capital market expectations followed by a review of various challenges and pitfalls that analysts may encounter in the forecasting process. The remainder of the reading focused on the use of macroeconomic analysis in setting expectations. The following are the main points covered in the reading:

• Capital market expectations are essential inputs for strategic as well as tactical asset allocation.

• The ultimate objective is a set of projections with which to make informed investment decisions, specifically asset allocation decisions.

• Undue emphasis should not be placed on the accuracy of projections for individual asset classes. Internal consistency across asset classes (cross-sectional consistency) and over various time horizons (intertemporal consistency) are far more important objectives.

• The process of capital market expectations setting involves the following steps:

  1. Specify the set of expectations that are needed, including the time horizon(s) to which they apply.

  2. Research the historical record.

  3. Specify the method(s) and/or model(s) that will be used and their information requirements.

  4. Determine the best sources for information needs.

  5. Interpret the current investment environment using the selected data and methods, applying experience and judgment.

  6. Provide the set of expectations and document the conclusions.

  7. Monitor outcomes, compare to forecasts, and provide feedback.

• Among the challenges in setting capital market expectations are:

  • limitations of economic data including lack of timeliness as well as changing definitions and calculations;

  • data measurement errors and biases including transcription errors, survivorship bias, and appraisal (smoothed) data;

  • limitations of historical estimates including lack of precision, nonstationarity, asynchronous observations, and distributional considerations such as fat tails and skewness;

  • ex post risk as a biased risk measure such as when historical returns reflect expectations of a low-probability catastrophe that did not occur or capture a low-probability event that did happen to occur;

  • bias in methods including data-mining and time-period biases;

  • failure to account for conditioning information;

  • misinterpretation of correlations;

  • psychological biases including anchoring, status quo, confirmation, overconfidence, prudence, and availability biases.

  • model uncertainty.

• Losing sight of the connection between investment outcomes and the economy is a fundamental, and potentially costly, mistake in setting capital market expectations.

• Some growth trend changes are driven by slowly evolving and easily observable factors that are easy to forecast. Trend changes arising from exogenous shocks are impossible to forecast and difficult to identify, assess, and quantify until the change is well established.

• Among the most important sources of shocks are policy changes, new products and technologies, geopolitics, natural disasters, natural resources/critical inputs, and financial crises.

• An economy’s aggregate trend growth rate reflects growth in labor inputs and growth in labor productivity. Extrapolating past trends in these components can provide a reasonable initial estimate of the future growth trend, which can be adjusted based on observable information. Less developed economies may require more significant adjustments because they are likely to be undergoing more rapid structural changes.

• The average level of real (nominal) default-free bond yields is linked to the trend rate of real (nominal) growth. The trend rate of growth provides an important anchor for estimating bond returns over horizons long enough for this reversion to prevail over cyclical and short-term forces.

• The trend growth rate provides an anchor for long-run equity appreciation. In the very long run, the aggregate value of equity must grow at a rate very close to the rate of GDP growth.

• There are three main approaches to economic forecasting:

  • Econometric models: structural and reduced-form statistical models of key variables generate quantitative estimates, impose discipline on forecasts, may be robust enough to approximate reality, and can readily forecast the impact of exogenous variables or shocks. However, they tend to be complex, time-consuming to formulate, and potentially mis-specified, and they rarely forecast turning points well.

  • Indicators: variables that lead, lag, or coincide with turns in the economy. This approach is the simplest, requiring only a limited number of published statistics. It can generate false signals, however, and is vulnerable to revisions that may overfit past data at the expense of the reliability of out-of-sample forecasts.

  • Checklist(s): subjective integration of information deemed relevant by the analyst. This approach is the most flexible but also the most subjective. It readily adapts to a changing environment, but ongoing collection and assessment of information make it time-consuming and also limit the depth and consistency of the analysis.

• The business cycle is the result of many intermediate frequency cycles that jointly generate most of the variation in aggregate economic activity. This explains why historical business cycles have varied in both duration and intensity and why it is difficult to project turning points in real time.

• The business cycle reflects decisions that (a) are made based on imperfect information and/or analysis with the expectation of future benefits, (b) require significant current resources and/or time to implement, and (c) are difficult and/or costly to reverse. Such decisions are, broadly defined, investment decisions.

• A typical business cycle has a number of phases. We split the cycle into five phases with the following capital market implications:

  • Initial Recovery. Short-term interest rates and bond yields are low. Bond yields are likely to bottom. Stock markets may rise strongly. Cyclical/riskier assets such as small stocks, high-yield bonds, and emerging market securities perform well.

  • Early Expansion. Short rates are moving up. Longer-maturity bond yields are stable or rising slightly. Stocks are trending up.

  • Late Expansion. Interest rates rise, and the yield curve flattens. Stock markets often rise but may be volatile. Cyclical assets may underperform while inflation hedges outperform.

  • Slowdown. Short-term interest rates are at or nearing a peak. Government bond yields peak and may then decline sharply. The yield curve may invert. Credit spreads widen, especially for weaker credits. Stocks may fall. Interest-sensitive stocks and “quality” stocks with stable earnings perform best.

  • Contraction. Interest rates and bond yields drop. The yield curve steepens. The stock market drops initially but usually starts to rise well before the recovery emerges. Credit spreads widen and remain elevated until clear signs of a cycle trough emerge.

• At least three factors complicate translation of business cycle information into capital market expectations and profitable investment decisions. First, the phases of the cycle vary in length and amplitude. Second, it is not always easy to distinguish between cyclical forces and secular forces acting on the economy and the markets. Third, how, when, and by how much the markets respond to the business cycle is as uncertain as the cycle itself—perhaps more so.

• Business cycle information is likely to be most reliable/valuable in setting capital market expectations over horizons within the range of likely expansion and contraction phases. Transitory developments cloud shorter-term forecasts, whereas significantly longer horizons likely cover portions of multiple cycle phases. Information about the current cyclical state of the economy has no predictive value over very long horizons.

• Monetary policy is often used as a mechanism for intervention in the business cycle. This mechanism is inherent in the mandates of most central banks to maintain price stability and/or growth consistent with potential.

• Monetary policy aims to be countercyclical, but the ability to fine-tune the economy is limited and policy measures may exacerbate rather than moderate the business cycle. This risk is greatest at the top of the cycle when the central bank may overestimate the economy’s momentum and/or underestimate the potency of restrictive policies.

• Fiscal policy—government spending and taxation—can be used to counteract cyclical fluctuations in the economy. Aside from extreme situations, however, fiscal policy typically addresses objectives other than regulating short-term growth. So-called automatic stabilizers do play an important role in mitigating cyclical fluctuations.

• The Taylor Rule is a useful tool for assessing a central bank’s stance and for predicting how that stance is likely to evolve.

• The expectation that central banks could not implement negative policy rates proved to be unfounded in the aftermath of the 2007–2009 global financial crisis. Because major central banks combined negative policy rates with other extraordinary measures (notably quantitative easing), however, the effectiveness of the negative rate policy is unclear. The effectiveness of quantitative easing is also unclear.

• Negative interest rates, and the environment that gives rises to them, make the task of setting capital market expectations even more complex. Among the issues that arise are the following:

  • It is difficult to justify negative rates as a “risk-free rate” to which risk premiums can be added to establish long-term “equilibrium” asset class returns.

  • Historical data and quantitative models are even less likely to be reliable.

  • Market relationships (e.g., the yield curve) are likely to be distorted by other concurrent policy measures.

• The mix of monetary and fiscal policies has its most apparent effect on the average level of interest rates and inflation. Persistently loose (tight) fiscal policy increases (reduces) the average level of real interest rates. Persistently loose (tight) monetary policy increases (reduces) the average levels of actual and expected inflation. The impact on nominal rates is ambiguous if one policy is persistently tight and the other persistently loose.

• Changes in the slope of the yield curve are driven primarily by the evolution of short rate expectations, which are driven mainly by the business cycle and policies. The slope of the curve may also be affected by debt management.

• The slope of the yield curve is useful as a predictor of economic growth and as an indicator of where the economy is in the business cycle.

• Macroeconomic linkages between countries are expressed through their respective current and capital accounts.

• There are four primary mechanisms by which the current and capital accounts are kept in balance: changes in income (GDP), relative prices, interest rates and asset prices, and exchange rates.

• In the short run, interest rates, exchange rates, and financial asset prices must adjust to keep the capital account in balance with the more slowly evolving current account. The current account, in conjunction with real output and the relative prices of goods and services, tends to reflect secular trends and the pace of the business cycle.

• Interest rates and currency exchange rates are inextricably linked. This relationship is evident in the fact that a country cannot simultaneously allow unfettered capital flows, maintain a fixed exchange rate, and pursue an independent monetary policy.

• Two countries will share a default-free yield curve if (and only if) there is perfect capital mobility and the exchange rate is credibility fixed forever. It is the lack of credibly fixed exchange rates that allows (default-free) yield curves, and hence bond returns, to be less than perfectly correlated across markets.

• With floating exchange rates, the link between interest rates and exchange rates is primarily expectational. To equalize risk-adjusted expected returns across markets, interest rates must be higher (lower) in a currency that is expected to depreciate (appreciate). This dynamic can lead to the exchange rate “overshooting” in one direction to generate the expectation of movement in the opposite direction.

• An investor cares about the real return that he or she expects to earn in his or her own currency. In terms of a foreign asset, what matters is the nominal return and the change in the exchange rate.

• Although real interest rates around the world need not be equal, they are linked through the requirement that global savings must always equal global investment. Hence, they will tend to move together.