<Portfolio DeRisk March-April 2017>
Part 4: Investment Outlook
Here we make the somewhat contradictory comment that while the probability of recession is the near-term is fairly low, the risk of a recession is significant. That is to say, while the chance of falling into recession within the next 12 months may be less than 30%, if we do it's almost inevitable we will see a painful correction in asset prices, reflecting both the lowering of return expectations, increase in risk and tighter controls over credit. Of course the existence of this risk isn't what we are (or should) be concerned about. Rather, we must be mindful of the potential impact and evaluate whether, on balance, we are being appropriately compensated for risk.
In a practical sense this calls for a significant amount of judgement. Different investors have different tolerances - and therefore require differing levels of compensation - for volatility, liquidity and time. Despite this, we believe that using a "return-on-risk" (or Risk Margin) framework is not only useful as a means of predicting long-term returns, but also as a way to forewarn investors of future price corrections.
This is well illustrated by Fig. 3. Here we compare the Implied Risk Margin to the returns actually achieved over the next 5 years. Looking to the far left of the chart we see that in the aftermath of the '87 crash Risk Margins got as high as 7% before steadily tracking lower for a bit over 10 years, getting below 2.5% by the time we hit the "dot.com" Bubble. Put another way, over the course of 12 years the price of earnings increased by around 180%, contributing about 9% p.a. to investor returns. Fig. XC7 : (XX* Show Components of Returns).
Although most people (with the benefit of hindsight) will accept that a risk premium of 6% or 7% is pretty generous, there is clearly no agreement on how small the premium can get before it no longer makes sense to be investing in equities.
<<bank interest margin>>
A general observation on market behaviour:
risk premiums become more valued as interest rates are lowered, and less valued when interest rates are higher. From a risk-returns perspective this is the wrong way around as risk premiums should reflect both current and future real earnings (we can obtain exposure to risk-free yield and inflation via fixed income instruments). Paying too much for earnings when interest rates are low essentially forfeits >>> miss out on potential monetary support // exposes us to interest rate headwinds. >>>
This "feels" right, but from a risk-return perspective
when interest rates are higher
with lower interest rates, risk premie
**When interest rates are higher >>
**Lower interest rates =>> makes risk premium
As interest rates are lowered, the
Over the past decade or so there has been much talk about the "new normal", with future equity returns likely to be lower. I feel, however, there is some confusion as to what this means for equity returns and the price we should be paying for earnings. Over the very long-term (30 years+) the bulk of equity returns are attributable to underlying earnings and earnings growth. As As a general statement, I agree that future returns from developed markets should be lower.
<< explain development of markets /. efficiency and productivity)
Here we offer some suggestions:
Bank interest margin
If you were less brave and sold out when the risk premium deteriorated to 4% you would have sold out in late 1995, only to see equity markets double in the years following before it's dramatic conclusion. The point here is that while measuring assets (and markets) according to Implied Risk Premiums helps identify market anomalies, it tells us nothing about when (or whether) the market will agree with you.
Fig. 4 provides us with another point of view regarding the influences of Risk Premiums: here we map Risk Premium on a shorter-term measure, the US 1 Year Treasury yield. Here it appears that markets (here we use the S&P500 as proxy) still contain a relatively healthy premium, at about 2.4%. What is important about this chart, however, is to recognise the relationship between changes in treasury yields and risk premiums. Statistical analysis of the YoY change in Treasuries versus Risk Premiums finds a correlation of -0.93%. With the Federal Reserve under pressure to raise rates sooner rather than later (to free up some "dry powder"), we have seen market expectations for target policy rates to hike from 0.98% (June 2017) to 1.84% (June 2019), and FOMC estimates of 3.0% by 2019. On current market valuations this would see the Risk Premium fall to between 0.9% and 1.6%.
Incidentally, in the above scenario, the cautious investor would have reinvested in mid-2002. For the period up to September 2006 the cautious investor would have earned an average return of about 9.5% versus 12%
while investing based upon an asset (or market's) Implied Risk Premium is sensible, it also provides very lit
(around September 2006). Meanwhile the cautious investor demanding his 4% risk premium would have reinvested in mid-2002, and by September 2006 would have a earned a return of 416%.
can prove not only useful, but effective as a means of alerting investors of times where markets
Despite these limitations, framing investment decisions from a return-on-risk (Risk Margin framework) helps identify times where assets are either greatly over (or under) valued.
Assuming we are comfortable with the idea erforming this analysis is relatively straightforward.
Despite this, we believe there are a number of general principles that most investors (most market participants) should be able to agree with:
(a) The return from any asset should be higher than that of a less risky asset;
(c) Each dollar of prospective earnings should be valued at less than a dollar;
Questioning whether we are being appropriately rewarded