Quantitative easing is designed to bring forward asset returns before we get the economic returns, in the hope that the confidence and increased wealth generate demand in the real economy. By its very nature its objective is to create a bubble in asset class valuations in the hope that the bubble drives the real economy forward
It is possible that markets have got well ahead of themselves, and could well be in bubble territory.
Why bubble territory? A bubble is when we have too much capital focussed on one asset class or area of economic activity, and where the valuation/capital allocated to those assets or areas far exceeds the ability of the economy to deliver the necessary return to justify the valuation and/or the allocation.
We can also have a bubble when a return has been brought forward in advance, or far too high an expected return discounted – this is your common garden stock market crash variety of bubble.
Quantitative easing is designed to bring forward asset returns before we get the economic returns, in the hope that the confidence and increased wealth generate demand in the real economy. By its very nature its objective is to create a bubble in asset class valuations in the hope that the bubble drives the real economy forward. Prior central bank moves, that is prior to 2008, were designed to impact the real economy, even though they ended up impacting valuation and allocation (debt), so we need to be concerned about policy directed unashamedly at asset values and their supply (housing).
If the necessary growth does not materialise, or is lower than is expected, or takes longer than expected, then we will most likely have a bubble of consequence for risk and return.
Bubbles are being engendered in stocks, bonds and in now in property as the Federal Reserve continues its policy of QE. These bubbles may not be as easily spotted as previously during periods of higher economic growth because we have failed to adjust prior rule of thumbs (our mental frames) to the boundaries and dynamics of the current paradigm.
In my previous post I pointed out the high share of national income that corporate profits currently have and suggested that valuation ratios need to adjust for this. Yes, corporate balance sheets in general are strong – interest rates have been low, governments have supported economic activity and labour costs have been cut –, but, consumers remain highly in debt, unemployment remains high, many economic areas (Europe and Japan) remain mired in slow to negative growth, governments worldwide are hitting constraints on expenditure and many key emerging markets, for example China, are coming up against real growth constraints that are only being allayed by booms in gross fixed capital investment, that are themselves likewise bubbles.
Moreover, we have increasing wealth inequality, that appears to have widened since the onset of the crisis, and therefore, for the benefit of overall economic growth, corporate earnings and asset prices cannot continue to maintain their trajectory. Historically, they never have, yet once again the market appears to believe that they can.
Just for the sake of analysis:
What if real economic growth/growth in national income were to average 1.5% (geometric) per annum over the next 10 years and that profits as a % of national income were to fall back to 8%. Note that government debt reduction would impact real economic growth and it would be difficult to see GDP moving ahead at its long term pre crisis clip.
What if there was no real growth in dividends over this period.
What if inflation averaged 2% over the period and 10 year bond yields were to average 2% per annum.
What if under one scenario markets adjusted immediately for the high profit margins and under another adjusted gradually over the 10 year period. Also, let us assume that under the former scenario we had a market and economic shock that saw a further temporary dip in earnings and a decline in market values of say 20%.
Under the gradual adjustment scenario, keeping valuation parameters the same, we would have negative real earnings growth of 1.5% per annum and a real dividend yield of 2%per annum. This is a real return of 0.5% per annum, or a risk premium of 0.5% above 10 year nominal bond yields of say 2%.
Under the immediate adjustment scenario, you would have real earnings growth of 1.5%, a dividend yield of 3.5% (of the starting value), or a real return of 5% per annum and a similar risk premium.
If we add back the temporary decline, equivalent to 1.84% per annum (1.2^1/10), representing the recapture of the 20% decline, the real return would be higher still.
Without the additional 20% decline, the return profile would be 1.5% capital return + 2.7% dividend = 4.2% per annum. The current P/E ratio would be 13.3. Obviously, as real earnings decline and capital gains increase by 1.5% per annum, in line with economic growth, the P/E ratio would rise. In other words, as profits as a % of national income falls, the discounted impact of the adjustment falls over time and the P/E rises.
If we expect P/E ratios to be lower at the end of the period than now, then real returns could well be lower. Also, if real economic growth were to average less than 1.5%, inflation were to rise above 2% and bond prices were to fall, real returns and risk premiums on risky assets could well be lower.
My concern is that many a return assumption that advisers/ors are using to model forward (in particular those based on Monte Carlo analysis) is just not going to pick up these risks.
Personally, if I was looking at modelling risk and return over time, I would be looking to cope with at the very least a real market return of 0.5% and building in an assessment of the impact of higher inflation and greater short term risks to economic growth.
While asset prices have moved ahead of the game and while much of the risk remains, it is over the next 6 to 9 months that we will possibly see the highest period of market risk. Just a hunch.
Oh and yes, in a low return/higher risk environment fees, expenses, transaction costs etc are going to weigh heavily. So your modelling also needs to build in the overall costs into the risk/return analysis.