Do you follow the long-term economic predictions published almost on a monthly basis by all sorts of renowned institutions? Do you read articles from world-renowned economists and analysts discussing global economic trends for many years ahead? Are you building complicated valuation models? And have you ever wondered if it actually makes sense at all?

If there is one thing analysts are notoriously bad at, it is precisely predicting the events that really matter. That is, inflection points (moments when the trend changes) and, as the name suggests, unexpected events - black swans.

Note that the vast majority of longer-term predictions and models are built in a way that expects the world to reach some sort of steady state at the end of the forecast horizon, where history has ended and all indicators will move at a constant rate to infinity. The correctness of this approach is shown by the idea of ​​a stock analyst who, at the end of 2016, values ​​ČEZ shares, expecting that the company will reach a stable state at the end of his five-year prediction in 2022 when nothing interesting will happen in the energy sector.

However, the world has never reached a stabilized state, despite the number of times it has gone beyond the horizon of all kinds of economic, econometric, stochastic, and other models. A fundamental influence on its development is precisely those black swans, which have the property of appearing unexpectedly and from different directions, turning all predictions completely on their head, moving company valuations and tens of percent, and pushing the achievement of that dream stabilized state to the next prediction horizon.

The most obvious example of a black swan in recent years is Covid-19. Before March 2020, the smartest economic minds spent thousands of hours making predictions for the coming years. They debated into the night whether assumption Y would have X or X + 0.1 impact on quantity Z and filled stacks of paper (or data stores) with their expert predictions. All this effort was then, literally overnight, managed to be invalidated by a random person somewhere in China who got a taste for bat soup.

After the shock of the arrival of Covid wore off, the world went back to normal and new predictions with new horizons began to form. What was their value? For the sake of interest, let's take a look at what kind of world we live in today according to the estimates of CNB analysts from last November (I'll leave the comparison with reality up to you):

• Real GDP grows by 5.3% year-on-year
• Inflation peaked at 7% in January 2022. In November, we reached last year's "monetary policy horizon" and year-on-year inflation hovers near the inflation target of around 3%, which confirmed the wise (and oft-repeated) words of Aleš Michl that inflation is only temporary and there is no need to keep raising interest rates.
• After a slight slowdown, the mortgage market bounced back with the two-week repo rate falling to 2.7% from its peak of 3.23% in the first quarter of 2022.
• Average real wages are growing at an annual rate of 1.9%, supporting household consumption.
• Whoever fixed energy prices in the winter lost, as the cost of natural gas rose just below the 100 EUR/MWh mark during January 2022, only to subsequently return to 40-50 EUR/MWh.
• You will pay a little over 20 crowns for a dollar.

Needless to say, this predicted world has definitely been transformed by another black swan in the form of a "special military operation" that was predicted in November 2021 by most analysts with high-quality connections to sources in the Kremlin.

Despite these fundamental problems, macroeconomic predictions, business plans, valuation models, budgets, and forecasts will always have a place in the world of finance (if only because everyone else relies on them and derives asset values ​​from them). It's just that when using and creating them, you need to approach the world with a probabilistic view and take into account that out of those (almost) eight billion people, sometimes someone will think that having bat soup for lunch tomorrow is actually a pretty good idea.