Karl Popper demonstrated in the Logic of Scientific Discovery that making predictions about the development of science is impossible and that consequently, since science is an important component of reality, any prediction about reality is equally impossible. So are the efforts of economists, analysts and strategists to predict the future of the markets meaningless? The question arises every year-end, when the demand and supply of forecasts reach their peak.
No one seems to care much, March 3 or August 24, what happens in the next 12 months, but between November and December, earlier and earlier every year, we stop to take stock and try to push our gaze into the future. We all know in our hearts that things will go the way they want, but the year-end exercise is not useless for those who produce it because it highlights any inconsistencies in their forecasts and it is not useless for those who use it provided they are clear the premises on which it was built.
In the big houses the assembly of the forecasts proceeds like an assembly line. Economists leave with their forecasts on growth, employment and inflation. Followed by monetary policy observers who, on the basis of what economists have said and considering the orientation of the central banks, predict the evolution of rates along the entire curve. Credit analysts add their forecast of credit spreads to the yield curve. A little aside, not always well integrated into the whole, currency analysts produce their forecasts on exchange rates.
At the bottom of the chain, equity strategists combine rate and foreign exchange estimates with bottom-up earnings estimates from analysts studying individual companies and produce a top-down estimate that is typically different from the sum of the individual bottom-up estimates because takes into account additional variables, such as market positioning for example. Political and geopolitical analysis, and fiscal policy forecasting in particular, should enter all stages of the process but this is sometimes not the case in practice because policy is implicitly considered impossible to model and predict.
Forecasters should never forget the limitations of this process. The first is linearity, i.e. the lack of feedback. The econometric models, for example, until 2008 did not include the trend of financial and real assets and considered them a simple effect of the macro variables, never a cause. Today, having seen the effects produced by markets on the economy during the Great Recession, this bias has been corrected, but still not fully.
The second limitation is psychological. The downstream analyst must start from the conclusions that the upstream analyst communicates to him even when he does not share them and is required to keep any dissent to himself. This turns him from a forecaster (it will be A then B) to a modeler (if A then B). In summary, whoever reads a forecast must make an effort to understand how many variables have been considered and how they have been related to each other. When there are few variables, the estimates can sound too perfect and precise, when there are many, however, the estimates tend to seem chaotic and lacking in internal logic.
For 2018 forecasts are easier than usual for the first part of the assembly line (economy and monetary policy) but are more difficult than in recent years for the last part, especially the stock market. The reason is that the first part is now being offered prefabricated by central banks. We thus know that global liquidity will reach an all-time high between the second and third quarters and then begin to decline very gently and that growth will still be good in 2018 but will be less brilliant in 2019.
We know that the Fed will raise rates at least three times and that the ECB will raise them in mid-2019, that the European curve will become steeper and the American one flatter, without flattening necessarily being the antechamber of the recession that some are talking about. On the stock exchanges, we know that profits will still rise quite a lot in the United States (if there is the tax reform, now in the pipeline) and not much in Europe, where much will depend on the euro. Japan is still doing well, some doubts about China.
By calculating only profits and rates, the stock exchanges will still be able to rise, albeit increasingly slowly, throughout 2018 and 2019 (Goldman's Kostin pushes optimism up to the whole of 2020). But there is a but. But this is not politics, which will continue to give us emotions in Spain, Germany, Italy and, in late 2018, the United States. The but are the internal factors of the market, its imbalance and its being vulnerable to an increase in volatility, its being increasingly concentrated on passive ETFs, perhaps leveraged, in the hands of investors with low risk appetite and ready to sell immediately in the event of a decline.
The more the markets rise, the more they go into rarefied atmospheres (even if legitimately driven by earnings), the more central banks withdraw excess liquidity, the more we all settle into a feeling of permanent calm, the more trading activity is delegated to robots and the higher the risk of flash crashes , i.e. of rebates that are as short as they are deep. Bitcoin, which is also in the midst of a majestic bull market and has excellent fundamentals on its side (not in the sense of intrinsic value, which cannot be understood where it is, but in that of the structural imbalance between supply and demand), is subject to periodically to terrifying flash crashes which, perhaps for a few seconds, reset its value.
Not all financial instruments are quoted in forms which provide for the suspension of trading beyond a certain low level. For example, limit down exists for index futures, but not for individual stocks. Then there are instruments that provide for automatic self-liquidation in the event of loss of value. From now on, therefore, it will not be enough to be right about the stock market trend, all in all still positive, and predict the moment for corrections (for Potts at the beginning of 2018, as sell the news on tax reform, for Hartnett in the second half of the year, as reaction to a first part without brakes).
It will also be necessary beware of sell stop loss orders and maybe, for those who regret having missed the train for the big upside, place some purchase orders on good and well-diversified securities with a price limit well below the current value.
