Integrating Earnings Forecasts, Valuation, and Return Prediction

This monograph addresses equity (enterprise) valuation and aims to provide practical tools for investors and financial analysts. We initially focus on an all-equity firm and adopt an approach to valuation similar to that carried out by fundamental analysts, attributing a market value of equity to the reported accounting values of equity, along with forecasts of future earnings and growth. It attempts to forecast future earnings, estimate intrinsic values, and derive a proxy for expected returns within a unified framework. It offers indicators of temporary mispricing and expected returns to investors in real-time.

A financial asset is expected to generate future cash flows. The asset’s present value is linked to the riskiness associated with the future cash flows in a specific manner in the capital markets. The concept of no-arbitrage plays a central role in this linkage. The dividend discount model (DDM) follows the no-arbitrage assumption in the capital markets. It is the fundamental pricing model; ultimately, all equity valuation models must reconcile with it.

In the clean surplus accounting framework, the dividend discount model transforms into an earnings-based valuation model, highlighting the critical roles of earnings and book value in equity valuation. However, earnings and book value are accounting attributes governed by accounting principles. This chapter addresses the impact of accounting conservatism on the dynamics of earnings information and equity valuation. Given the inherent difficulty of forecasting earnings over multiple periods and consistent with the analysis in the last chapter, we focus on predicting one-period-ahead earnings.

Earnings are part of sales revenue. But sales are generated from investment assets. From a shareholder’s perspective, earnings should at least equal the required return of equity capital on the initial investment. From a firm’s point of view, it should cover the capital charge, that is, the cost of equity capital on the book value of equity. Earnings can also be generated from the firm’s goodwill and other intangible assets. Future earnings can also be affected by conservative accounting principles. Although much progress has been made in understanding how factors affect earnings generation, less progress has been made in defining the functional dependence of earnings on these factors. In this chapter, we explore the determinants of future earnings.

Under the assumption that the aggregate capital market is efficient, we can use stock prices to replace the intrinsic values in Eqs. (4.​3), (4.​9) or (4.​12) to run cross-sectional or industry panel data regressions to forecast future earnings of individual firms. We can then use the forecast earnings to value the equity shares of individual firms and predict one-period ahead returns.

The accounting for operating activities differs from the accounting for financial activities. To an approximation, the accounting for financing activities is at fair value, whereas the accounting for operating activities usually uses the (amortized) cost basis. Additionally, the fundamental origins of risk must be rooted in a firm’s operations; most financing activities serve to attenuate operating risk, such as expanding operating activities by issuing more debt and adding leverage, or hedging operating risk by holding net cash and cash equivalents. In this chapter, we introduce a unified framework for forecasting operating earnings, valuing invested capital (or net operating assets), and determining the cost of capital at the enterprise level. Based on the intrinsic relationship between the value of invested capital and one-period ahead operating earnings, we can simultaneously estimate industry-level valuation parameters, including the weighted average cost of capital (WACC) and the expected long-term growth rate of operating earnings. This contrasts with conventional valuation analysis, which begins with forecasts of sales revenue. In an efficient market, macroeconomic conditions and industry trends are reflected in the current value of stocks, hence the value of invested capital. We can let the data reveal their intrinsic linkage, guiding the applications of machine learning in asset pricing. Using industry valuation multiples as proxies for individual firms’ valuation parameters, we can estimate synthetic valuation multiples to value an enterprise and evaluate an individual firm’s WACC.