Crowding out is back
The term "crowding out" was first used in the last century. It referred to a situation where the supply of capital did not increase significantly, but borrowing continued to rise, faster than the money supply was growing. As a result, interest rates rose until borrowing declined. Only the government was unaffected by the higher interest rates; it simply continued to lend at the same rate. This meant that the private sector could borrow less. Hence the expression "crowding out."
Something similar is happening now as well:
- Given the current phase of the economic cycle, the government deficit is far too large. To the extent that this drives up interest rates, it hardly leads to measures to reduce the deficit. Politicians are too divided to take such action.
- There is now massive investment in all things related to AI. More and more borrowing is required to finance this. However, here too, higher interest rates have little effect on the amount being borrowed. This is because the major tech companies that are currently investing so heavily do not believe the fears currently circulating in the markets, namely that AI could lead to overcapacity. These companies expect that, despite the massive investments they plan to make, there will still be a huge shortage of AI capacity after 2030. They therefore expect to generate high profits from their planned investments. A few percentage points higher interest rates do nothing to change that.
- Given the high government deficits, investments in all things AI, as well as high stock prices and low credit spreads, most economies are growing at a fairly rapid pace, especially the U.S. economy. Combined with sharply rising energy prices and a tight labor market, this is leading to excessively high inflation. That is why many central banks are tightening monetary policy. This means a slowing growth in the money supply.
When all these factors are combined, it means that Western economies, in order to keep inflation under control amid rising energy prices, must reduce growth to below potential growth. This must be achieved primarily through "crowding out" the business sector outside of AI and outside of companies that benefit from government spending. However, once these sectors are sufficiently affected to push the growth of the overall economy below potential growth, monetary policy must be loosened again, because, given the high levels of debt, a recession must be avoided.
So much for the theory. Now the question is how a central bank would implement this in practice. It essentially boils down to continuing to tighten monetary conditions until something breaks. For example, much weaker economic data, falling stock prices, or credit spreads widening too far. Furthermore, there must not be too many bankruptcies among non-AI companies; otherwise, the economy would be eroded too severely.
If we apply this to the current situation, the S&P 500 index, for example, remains very high, and the economies in the U.S. and Europe are still growing at a fairly rapid pace. However, if one looks more closely at the stock markets and credit spreads, cracks are beginning to become visible. Shares of companies outside the AI sector are increasingly experiencing sharp declines. For example, the Russell 2000 index has fallen much more than the S&P 500.
There is also a connection here with exchange rates:
- In the West, investment in AI is much higher in the U.S. than in Europe. This means the U.S. attracts much more foreign capital to finance this. Consequently, there is strong demand for dollars from this source.
- Due to the wave of investment in AI, the U.S. economy has become much less sensitive to higher interest rates. The Fed will therefore have to raise interest rates to fairly high levels to push growth below its potential. This is all the more true given that investments in AI and the large government deficit continue to stimulate the economy.
- At some point, however, this cycle will come to an end. The business sector, excluding the major tech companies, will then find itself in such a tight spot that there appears to be a risk of too many bankruptcies and a significant slowdown in economic growth. In the U.S. in particular, monetary policy will then need to be loosened. This could also happen if, for whatever reason, the stock prices of the major tech companies were to fall sharply (for example, because the market becomes concerned that overcapacity is now being created). This development would actually be negative for the dollar, and positive for the euro. The European currency is currently weak because Europe has virtually no major AI tech companies. However, this will become an advantage if fears of overcapacity arise and prices for AI usage decline. While Europe is not a major producer of AI, it is a major user of it.
