We don’t write very often on macroeconomic policy anymore, mostly because we are acutely aware of the complexity of economic systems. But it’s also because we are deeply upset at how we’ve permitted our policymakers to act over the past few years. While the Krugmans of the world loudly advocate mortgaging our childrens’ futures to preserve a bloated, over-indebted, short-termist, consumption focused status quo that channels virtually all of society’s wealth creation into the hands of the privileged few, we can’t help but think this will all end very badly. Worse, those who stand to lose the most – our children – are the least to blame for this predicament.
But I digress. This article is actually about how central banks have orchestrated a large overshoot in asset prices, and what this means for likely future returns to stocks and bonds in particular. We’ve written on this topic before, but this article approaches the problem from the point of view of central bank policy rather than analyzing valuation metrics.
The motivation for this article comes from a Bank of England “Quarterly Bulletin” entitled, “The United Kingdom’s quantitative easing policy: design, operation and impact“. We feel this bulletin has profound implications for investors, as it speaks to central banks’ goal of stimulating demand and inflation expectations primarily through the channel of ‘portfolio balance effects’: that is, inflating prices of stocks and bonds.
Specifically, the BOE believes that the primary impact of QE in the short term is to cause holders of QE eligible securities – primarily non-bank financial institutions such as insurance companies and pensions – to surrender these securities to central banks in exchange for newly printed money. The BOE then expects this cash to be redeployed toward higher yielding securities, such as credit and equities. This dynamic, compounded by effects from policy signalling (central bank jaw-boning) and liquidity enhancements, would serve to drive the real prices of these assets materially higher in the short term. The ultimate goal is to engender a ‘wealth effect’ whereby asset holders feel more confident and ratchet up spending commensurately, increasing end demand and inflation.
Figure 1. QE Transition Channels
Source: Bank of England
From the paper:
Portfolio balance effects: Central bank asset purchases, through this channel, push up the prices of the assets bought and also the prices of other assets. When the central bank purchases assets, the money holdings of the sellers are increased. Unless money is a perfect substitute for the assets sold, the sellers may attempt to rebalance their portfolios by buying other assets that are better substitutes.(3) This shifts the excess money balances to the sellers of those assets who may, in turn, attempt to rebalance their portfolios by buying further assets — and so on. This process will raise the prices of assets until the point where investors, in aggregate, are willing to hold the overall supplies of assets and money. Higher asset prices mean lower yields, and lower borrowing costs for firms and households, which acts to stimulate spending. In addition, higher asset prices stimulate spending by increasing the net wealth of asset holders. While policy signalling effects affect expected policy rates, portfolio balance effects work by reducing the spreads of longer-term interest rates over expected policy rates (term premia) and the required return on risky assets relative to risk-free assets (risk premia) more generally. ... in the impact phase, asset purchases change the composition of the portfolios held by the private sector, increasing holdings of broad money and decreasing those of medium and long-term gilts. But because gilts and money are imperfect substitutes, this creates an initial imbalance. As asset portfolios are rebalanced, asset prices are bid up until equilibrium in money and asset markets is restored. This is reinforced by the signalling channel and the other effects of asset purchases already discussed, which may also act to raise asset prices. Through lower borrowing costs and higher wealth, asset prices then raise demand, which acts to push up the consumer price level. ... The Bank of England’s asset purchase programme has attached particular importance to the portfolio balance channel. That is why purchases have been targeted towards long-term assets held by non-bank financial institutions, like insurers and pension funds, who may be encouraged to use the funds to invest in other, riskier assets like corporate bonds and equities. Before asset purchases began, the main holders of gilts were UK non-bank financial institutions and overseas investors. Gilts only represented a modest part of UK non-bank financial institutions’ overall portfolios, suggesting they might be prepared to reinvest some of the money from gilt sales in other assets. Overseas investors might be more inclined to choose to invest in foreign assets. However, to do so they would need to change their sterling for foreign currency, putting downward pressure on the exchange rate. And, since all central bank money has to be held by someone, those who received the sterling might then choose to invest in other sterling assets.
The “Impact Phase” of quantitative easing, as described in the above paragraphs, is illustrated in Figure 2 below. The red line shows the effect on real asset prices like stocks and credit markets, while the blue line shows the increase in the broad money supply as a function of QE. Note how asset prices track excess broad money creation, but with a meaningful amount of excess torque. Clearly we have seen this asset pricing dynamic play out around the world wherever central banks are aggressively pursuing quantitative easing, and the effect is most profound in regions with the most aggressive monetary policies, such as the U.S. and Japan.
Figure 2. The ‘Impact Phase’ of Quantitative Easing (red line: real asset prices | blue line: broad money)
Source: Bank of England
You may be thinking, ‘So far, so good. What’s the problem?” Here is where things get interesting. Again directly from the paper:
In the adjustment phase, rising consumer and asset prices raise the demand for money balances and the supply of long-term assets. So the prior imbalance in money and asset markets shrinks, and real asset prices begin to fall back. The boost to demand therefore diminishes and the price level [inflation] continues to increase but by smaller amounts. The whole process continues until the price level has risen sufficiently to restore real money balances, real asset prices and real output to their equilibrium levels.
The language in the paper is predictably benign, so as not to alarm readers about the implications. But the benign language is misleading, especially with respect to the impact of attenuating QE on real asset prices like stocks, bonds and real estate. Essentially, portfolio balance effects are driven by the flow of newly printed money which is used to purchase government bonds primarily from insurance companies, wealth funds and pensions. While bond purchases – QE – continues to take place, these institutions are continually faced with holding newly minted low-yielding cash assets, which they then roll out the risk curve in search for higher yields. As assets are priced ‘at the margin’, these marginal dollars serve to continually apply upward pressure to prices. No doubt this engenders other feedback mechanisms related to (over)confidence, leverage effects, and momentum which carries markets far beyond equilibrium. However, when this positive feedback mechanism ends, “real asset prices begin to fall back“. Let’s see what that looks like according to the Bank of England in Figure 3.
Figure 3. The ‘Adjustment Phase’ after the end of Quantitative Easing
Source: Bank of England
It’s clear from Figure 3. that the Bank of England believes the mechanism of action for QE to result in massive short-term asset price inflation during the ‘Impact Phase’ of Quantitative Easing. However, when asset purchases slow or stop during the ‘Adjustment Phase’, they fully anticipate real asset prices to revert to prior equilibrium valuations. If we apply this concept to equity markets, we would expect markets to revert to long-term average valuations which, depending on the valuation method implies a real price drop of between 50% and 80%, per Figure 4. below.
Further, markets rarely just revert to ‘equilibrium’; just as markets clearly overshoot to the upside from leverage effects, herding and other feedback mechanisms, they are equally likely to overshoot to the downside. Indeed, in order to preserve a stable equilibrium, market must definitionally trace out the same magnitude of time and value below equilibrium as they do above it.
Figure 4. Current valuation versus long-term equilibrium
It’s also worth noting that central banks have a very poor track record of managing this ‘Adjustment Phase’ effectively, as we have seen repeatedly in Japan over the past two decades or more, and as we saw in the U.S. and Europe in 1937 when policymakers unwound stimulative policies after the Great Depression. Figure 5., which we captured from a recent Bridgewater presentation illustrates how monetary tightening periods have negatively impacted asset prices at various periods in the past.
Figure 5. Drawdowns for traditional 60/40 endowment model portfolio and Bridgewater’s ‘All Weather’ risk parity based strategy during monetary tightening periods
Speaking of Bridgewater, Ray Dalio was on TV yesterday claiming that investors should expect a nominal 4% in total returns from equity markets over the next 10 years. That’s a little bit more ‘balanced’ than our own statistical forecast of about 0% real, or 2.5-3% nominal returns over the same period, but definitely in the same ballpark. Figure 6. from the same presentation as Figure 5. shows how Bridgewater may have constructed its forecast. Returns since 1970 on a traditional 60/40 endowment style portfolio have averaged about 9.9%. This return can be decomposed into the return on cash plus the excess return on risky assets in the asset allocation. Note that over the full period cash returned 5.6% nominal while the risk premium on the asset allocation amounted to just 4.3%.
If we fast-forward to today, we note that current cash yields (for 5 year money) are 1.4%, about 4% lower than the average yield over the past 43 years. The question is, what should we assume in terms of excess returns on a traditional asset allocation in order to arrive at our total return estimate? I guess Bridgewater expects about 3%…
To reiterate what we’ve said many times in the past, while long term performance – 7 to 20 years – may look pretty bleak, anything could happen in the short term. All our research suggests markets are nearly impossible to forecast in the intermediate term (several months to 5 years or so), and markets have a history of moving well beyond equilibrium once momentum takes hold, but over the long-term we expect gravity to exert its inexorable pull. Looks like the BOE agrees.