We conduct a novel systematic textual analysis of the discussion in the quarterly reports of China fund managers, from which we infer their near-term expectations for Chinese monetary policy. We show that this aggregate index of manager expectations performs well as a forecast of Chinese monetary policy, that fund managers act on these expectations, and that correctly anticipating shifts in policy improves fund performance; thus, manager skill is an important determinant of fund returns.
The adoption of unconventional monetary policies over the past decades has made monetary policy much more complicated to understand, thereby increasing the importance of central bank communications and transparency. This makes it important to ask, how does monetary policy affect the asset management industry? Ammer et al. (2020) evaluate implications of this policy communications situation for China fund managers, and hence ultimately for Chinese monetary policy itself. Our starting point is a novel, systematic textual analysis of the qualitative discussion in the quarterly reports of China’s mutual fund managers, which gives rise to a large panel data set (4,571 funds, between 2008 Q3 and 2019 Q1) of mutual fund managers’ near-term expectations about shifts in Peoples Bank of China (PBoC) monetary policy. Each data point is a value between −1 and 1. The expectation measure is positive if a manager expects a tightening of monetary policy, negative if the manager expects an easing, and zero if the manager expects no change. The absolute value reflects the level of certainty and expected magnitude of the monetary policy change. For instance, the expectation measure for “We believe that monetary policy will continue to ease in the second quarter,” a statement taken from Penghua State-owned Corporate Bond Fund’s 2015 Q1 report, is −1.
Note: The solid line is the consensus forecast, measured as the mean of forecast scores across managers. The shaded columns are the monetary policy index constructed with the required reserve ratio, the benchmark deposit interest rate, and the benchmark lending interest rate. The darkly shaded columns indicate periods of tightening. The lightly shaded columns indicate easing periods. Monetary policy is unchanged in periods with no columns.
Monetary policy expectations and portfolio adjustment
We examine how monetary policy expectations affect financial market participants’ investment decisions by matching the constructed expectation data to mutual fund managers’ investment history. We document that managers of money market funds buy (sell) long-term assets when expecting an easing (tightening) of monetary policy (Figure 2), consistent with the prediction of standard models. This finding provides direct evidence that monetary policy expectations serve as an important factor behind investors’ portfolio choice. It implies that systematic revisions of monetary policy expectations among market participants can induce significant rebalancing of asset holdings at the aggregate sector level [Feroli et al. (2014)], and thus potentially act as a channel of monetary policy transmission.
Note: The dashed curve displays the money market fund consensus forecast measured as the mean of forecast score across money market fund managers. The solid curve is the average weight of short-term asset holding weighted by fund size. Short-term asset is defined as an asset with a maturity less than 60 days. The shaded columns are the monetary policy index.
Characteristics of superior forecasters
While the consensus forecast index is validated as an exceptional predictor of monetary policy, there is substantial heterogeneity of forecast skill, measured by how often managers correctly anticipate shifts in monetary policy. An interesting question is which mutual fund managers best forecast shifts in monetary policy? We find that higher performers tend to manage a larger fund and charge higher management fees, both of which reflect how the market judges their skill. Interestingly, good forecasters also have more years of experience in the asset management industry and are more likely to hold a Ph.D. Both results are consistent with the idea that acquiring the necessary information to optimize investment decisions entails nonnegligible resource input, as first formalized by Grossman and Stiglitz (1980). Finally, Beijing-based fund managers are better forecasters of monetary policy, consistent with managers obtaining an informational advantage from geographical proximity to policy makers.
The effect of interest rate liberalization on mutual fundsA crucial change in China’s monetary policy practice in the last two decades is interest rate liberalization, which was a process of the PBoC’s transition from imposing interest rates through administrative orders to influencing market rates by managing its own balance sheet. While previous work has shown that interest rate liberalization has a significant effect on the transmission of monetary policy to the real economy [Liu et al. (2020)], we examine how it affects the mutual fund industry. We find that correct predictions of monetary policy improved money market funds' performance in the period prior to 2013, when market interest rates (such as inter-bank rates), which largely determine money market fund returns, remained in a narrow range around the benchmark deposit rate set by the central bank. After 2013, market interest rates have been less closely connected to benchmark policy interest rates, at least in the short run (Figure 3). Hence, correctly predicting the benchmark policy interest rate did not help the money market funds achieve a superior return in the latter part of the sample. However, it turns out that bond funds typically earn higher returns when their managers correctly predict the near-term direction of monetary policy, regardless of the policy regime. A possible interpretation of this finding is that bond funds invest in relatively long-term assets, whose value is heavily influenced by both the level of policy interest rates and revisions to near-term expectations. In contrast, correct predictions of monetary policy do not improve fund performance for equity funds and mixed funds, consistent with factors other than monetary policy being more important drivers of stock prices.
Note: The blue line is the benchmark three-month deposit rate. The red line is the weighted average of three-month inter-bank offered rates. The green line is the weighted average of money market funds' return. The patched area indicates post-June 2013, after which the market interest rates are allowed to float in a wider range around the benchmark policy rates.
Monetary policy expectations and fund flows
An important question about monetary policy is whether expected shifts induce systematic fund flows. We find significant net inflows to Chinese money market funds associated with near-term prospects for monetary policy easings (Table 1). This is consistent with strategic substitution by yield chasing depositors from bank deposits to money fund shares. Specifically, an easing (tightening) of monetary policy typically widens (narrows) the interest spread between the deposit rate and the money market funds’ yield, due to the loose link between the wholesale bank funding market and the short-term securities market in China. Interestingly, the sign of this relationship is at odds with what is found for banks in most other countries, where the link between wholesale bank funding and the short-term securities markets tends to be much closer, and low rates tend to depress net interest margins [Claessens et al. (2018)]. All else being equal, this substitution mechanism could weaken transmission of Chinese monetary policy to the real economy, because most bank borrowers do not have access to money-market financing. In contrast, there is no significant substitution between bank deposits and bond and equity investments. A possible reason is that bond and equity investments are subject to trading costs, while the trading costs for money market funds are close to zero.
Note: The dependent variables are the net inflow rates to each fund type. Et (mpt+1) is the consensus forecast. Rt is the weighted average of three-month inter-bank offered rates. Data is quarterly from 2008 Q3 to 2019 Q1. *, **, and *** denote significance at the 10%, 5%, and 1% levels, respectively.
(John Ammer is a Senior Economic Project Manager in the International Finance Division at the Federal Reserve Board in Washington, DC; John Rogers is a Senior Adviser in the International Finance Division of the Federal Reserve Board; Gang Wang is an assistant professor of at Shanghai University of Finance and Economics; Yang Yu is an assistant professor of economics at Shanghai University of Finance and Economics.)
Ammer, J, J. Rogers, G. Wang, and Y. Yu (2020): “Monetary Policy Expectations, Fund Managers, and Fund Returns: Evidence from China,” Federal Reserve Board, International Finance Discussion Paper 1285, www.federalreserve.gov/econres/ifdp/files/ifdp1285.pdf