La prévision est un art difficile… – Valérie Gastaldy, DayByDay

… surtout quand il s’agit de l’avenir ! Valérie Gastaldy nous explique très précisément qu’elle ne prédit pas l’avenir des marchés. En revanche, elle sait analyser la déformation de certains paramètres ou d’informations, susceptibles de faire évoluer l’état d’esprit des investisseurs et donc impacter les marchés. Alors pour 2020…

Why should the US economy improve next year ?

The economy evolves along regular cycles. We study those cycles thanks to long term data, mainly in the US economy. Our approach differs from the NBER approach. The NBER submitted a definition ex-post in order to fit the data that regularly displayed quarters of GDP contraction.  Our definition of cycles is ex-ante.

We assume that all living organisms follow patterns of expansion and contraction. Markets are crowds of investors, and crowds, groups,  have a life of their own. They are borne, they expand, and they die around common beliefs that get  adopted, shared and finally replaced by other beliefs. In turn, market crowds will adopt various macro-economic stories. The bulls today think that money supply is plentiful and the key driver. The bears have underlined the slowing economy and the unresolved reasons of that slowdown. As the S&P 500 prices are showing the dominance of the bulls, bearish investors are, little by little, moving to the bullish group. They change their opinion because the average human being is not able to go against the majority for a long time. If you are walking North and cross one person walking South, you keep on walking North. If you meet ten persons walking South, you will wander where they are going, and what is interesting South. But you will go on walking North. Everyone has a different tipping point (according to his conviction and self-confidence), but in the end, we will all assume that those who walk South have an important piece of information that we ignore, and we will turn to follow them. The North-walking crowd will have disappeared.

In the economic environment, the bullish and the bearish crowds share the same information, but they weigh it differently. Our view is that those weighs are absolutely subjective, though of course influenced by data. Those weighs are attributed according to the confidence : confidence in the models, confidence in the political environment etc. It has been proven that experts in general, and economic experts, are very poor at calibrating their forecasts. Behavioural biases influence them.

This is why we think that forecasting future events is not what is necessary (we doubt it is possible!), but forecasting how human behaviour and confidence will evolve -and therefore influence markets and the economy- is the best key to making decisions in that uncertain environment. Our research on the Industrial Production and the ISM Manufacturing shows that the US economy should improve next year. Of course, this may be caused by the recent improvement of the Chinese-American relations and global trade. But we do not know … the future !  In the past, when human groups have been living equivalent economic ups-and-downs, their reactions, their decisions, have had consequences. The facts were different, their actions were different, but those actions occurred at a time that is pretty regular, and the consequences come after a delay that also is regular. It is easy to predict when people will feel hungry if you know when they had their last meal and what is their past meal-schedule. It is a little more difficult to forecast the economic rhythm, but we rely on the same mechanism. Living organisms are predictable, not events.

Source : DayByDay

The chart above displays the average Industrial Production Berry cycle ( lasting 33 to 35 years ) our research has identified, based on data since 1885. That cycle is black. The red line is the current cycle, which started in December 2012.

Berry cycles extend over three Juglar cycles, and they start with a short transition called Set-up. Each of the following three cycles (between vertical dotted blue arrows) has its own characteristics, and its own role in the overall development of the economy. 

The first Juglar -which we are witnessing right now- transitions to a new model that will develop during the second cycle in order to die in the third cycle.

That first cycle is hesitant. It is exploratory in the economy, and in the society. It prepares the ground for either new technologies to boost productivity gains, or important social changes that will re-organise society. Those changes will only become obvious in the second Juglar cycle. They will allow a real development, and the 12-month variation of Industrial Production during that second cycle is almost continuously above zero. 

The 12-month variation of Industrial Production is a lot more hesitant during the first cycle. You can see that it drops several time below zero. Periods of expansion are short-lived, on average. 

We have noticed that in the XXI century, variations have been less and less pronounced, be it the expansions or the recessions -except 2008/2009. The NBER sees only one economic cycle since 2009, because there has been no recession according to its strict and ill-founded definition. In our view, even though the amplitude of the cycle has been restrained, all the beats are still present and occurring on time -with the precision we can expect from our methodology. The Industrial Production cycle has been through all normal ups and downs, it has been « breathing » normally. The « cycle » is not abnormal to us.  The chart is clear. The Industrial cycle will turn up soon.

Source : DayByDAy

The chart above displays the average Industrial Production Juglar cycle (lasting 10 to 11 years ) our research has identified in black. It is the detailed evolution of the three Juglar cycles that compose the Berry cycle above.

 The red line is the current cycle, which started in December 2015.

From late 2015 to now, the cycle has been very slow. Its expansion has never been spectacular – it remains well below average- but its 12-month variation has barely dropped below zero. It is now in the perfect situation to recover. 

We nonetheless have to be careful as we are still in the first stages of the cycle. When we build this chart, we have to make an assumption about the length of the current cycle. We are assuming that it will last 10.7 years, but it could just as well  11.3. This would delay the bottom by a few months. 

Very probably, we can say that the bottom in US Industrial production will come by the first two months of 2020.

Source : DayByDAy

The chart above displays the average Juglar Manufacturing ISM cycle ( lasting 10 to 11 years ) our research has identified, based on data since 1948. That cycle is black. The red line is the current cycle, which started in October 2011.

The ISM cycle has fallen heavily in 2019. It is taking a little longer than we thought to recover, but the tendency to optimism is already being felt, as we had announced. As we explained already a few times, cycles are based on human behaviour. They predict how humans will adapt to change, they do not predict the future of events. 

We have adjusted the length of our cycle in order to match what happened recently at best. We know that we have to go through a pretty clear recovery soon. For some three to four months early next year, the Manufacturing ISM should jump. It will not improve to a very high level as what will happen in 2020 Q2 to Q3 will be a decreasing improvement (the Rate-of-Change will decrease but remain positive).

The beginning of the year should be the most enthusiastic period for this indicator.

Partagez cet article.

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)
Valérie Gastaldy
A propos de Valérie Gastaldy 3 Articles
Valérie Gastaldy, MSTA et CFTe, est stratégiste et analyste Global Macro et Equities, au sein du bureau d’analyse DayByDay. Diplômée de l’ESSEC, elle a obtenu son diplôme d’analyse technique internationale, il y a 20 ans. Elle a débuté sa carrière en tant que market maker sur options, pour devenir pas la suite trader. Elle a co-fondé DayByDay. Elle est à présent en charge des indices internationaux et elle gère l’allocation globale d’actifs pour des gestions alternatives et long only. Valérie participe également à l’allocation tactique pour les actions. Enfin, elle co-anime le programme de formation DayByDay.

Soyez le premier à commenter

Laisser un commentaire

Votre adresse de messagerie ne sera pas publiée.