Back in my youth I had fond memories of visiting my aunt and uncle in the suburbs of Washington DC. She was my father’s sister and married to a “yank” in London just after the war. That “yank”, or who I was privileged to know as Uncle Col. Leroy Tidler, came from Virginia in the Blue Ridge Mountains. On one occasion we took a drive to the Luray Caverns, which I fell in love with. Little did I know that the drive to the caverns took us close to Ashburn/Loudoun County – an area now referred to as “Data Centre Alley”.  In fact, Virginia is the number one state in terms of data centres, with over one gigawatt (GW) under construction and another 6 GW in the pipeline.

Back in November, we published a newsletter entitled When Crowded-in becomes crowded out.  In Economics 101, this refers to government borrowing soaking up all the available cash in the room, leaving businesses with scraps. In that newsletter we added AI-related capex to the list of potential sources of capital displacement. While the focus then was on crowding out of non-AI business investment in plants and equipment, this week I want to take a look at how this can also be impacting residential construction. Given that housing affordability has become a lightning rod this U.S. mid-term election year, increasing the supply of homes is critical. Failure to do so has implications for the state of the American consumer, the economy and ultimately the future direction of stocks.

The above chart shows U.S. housing starts and spending on data centres. This is a sub-group of non-residential office construction and is relatively new. While residential housing starts in the US have stagnated, spending on data center construction has effectively gone vertical. We are seeing a massive reallocation of real-world resources—copper, cement, and crucially, skilled labor—away from residential construction and toward the build-out of AI data centers. It turns out that the same electrician you need to wire a new subdivision is currently being paid double to wire a server farm for a chatbot. The exodus of foreign construction workers from the site has compounded this problem – yet another example of externalidad negativa no comprendida (negative externality not understood).

This week, the Federal Reserve decided to leave interest rates unchanged (first pause since last summer) and this was fully anticipated and reasoned by everyone outside of the oval office. In the Fed’s eyes, the economy appears to be okay, and the job market might be stabilizing (tell that to the 16,000 workers that are being let go at Amazon), however, inflation is looking a little elevated. Definitely not the recipe for cuts to interest rates, yet even if Trump could perform a jedi mind trick on Powell and get further reductions, it is unclear how this would cure the housing affordability problem. True, lower borrowing costs helps the household budget, but it also stimulates housing demand. With constrained supply, home prices may not go down but up instead.

Speaking of squeezed consumers, we have to talk about the other side of the coin: the stock market. Even with short-term interest rates on hold, financial conditions are actually loosening because the stock market is acting like a runaway train. On Wednesday morning, the S&P 500 broke briefly above the 7,000 level to set a new record high and, in so doing, has added to a massive "wealth effect." In other words, when portfolios are up, people feel richer. They spend more on large-ticket items, like homes, and because supply is restricted, prices are bid up and affordability slips further away. The problem is that while wealth in the upper section of the “K-shaped economy” is concentrated in portfolios and real estate, the costs for the lower section (housing, food, insurance) are not going down.

To make matters more complicated, the US dollar has slid to a 4-year low this week. For the average American household, this is the "sting in the tail." A weaker dollar means imported goods—the stuff filling the shelves at Walmart and Amazon—get more expensive. So, you have a consumer who is already feeling the pinch of housing affordability (thanks to the data center crowding out) now facing potential price hikes on everyday goods.

Without a wealth effect, the average consumer is likely to feel poorer and perhaps start to behave as so. Case in point, the Conference Board Consumer Confidence Index fell to 84.5 in January which is the lowest we’ve seen since 2014. Now there are more factors at play here than just housing affordability, especially given geopolitical developments since the start of the year; but the main point is that people aren’t happy. That said, they aren’t behaving that way at the cash register so far. Since May of last year, there hasn’t been a monthly decline in retail sales. When consumption accounts for close to 70% of total US economic demand (GDP), that is great news. The question is what happens if the spending taps are shut off?

For investors, the bigger question is whether the apparent disconnect between stock market exuberance and underlying cracks in the economy is something to worry about? Let’s go back to the consumer confidence data for a moment. How many times have we seen the stock market at or near record highs while consumer sentiment was this depressed? The answer, at least in modern times, is never. The above chart illustrates the S&P 500 against the consumer confidence index. In 2000 (the “dotcom” peak) stocks were at records, but consumers were euphoric (the confidence index hitting highs of 111.4). In 2007 (pre-Great Financial Crisis), stocks had just hit new record highs, and sentiment was a healthy 90.0 plus.

Historically, the stock market and the consumer eventually have to agree. Either the consumer snaps out of their funk and spends enough to justify equity valuations (the soft-landing scenario), or the market wakes up to the reality that the engine of the economy—the consumer—is out of gas. Wall Street economists are pointing to the AI productivity boom as a reason to ignore the gloomy consumer. They argue "this time is different." At Pyle Wealth Advisory, we know that the four most dangerous words in investing are "this time is different." So, while we remain invested, we also are on the lookout for signs the consumer is throwing in the towel.

When I look back, my first memories of Virginia and DC were the summer of 1971. The S&P500 had just rebounded from the sharp loss of 1969-70 and was trading around 100. The next nine years would see a lot of volatility and by the end of 1979 the index closed at 108. Not a great decade indeed. Leroy passed in 1980 and was buried in Arlington. Even though he was an eternal optimist, I don’t think he would have believed that the S&P would break through 1000, let alone 7000. But, if he were here, I think he might say “watch out for breaking stalagmites”.

On behalf of the Pyle Wealth Advisory team, have a wonderful weekend.     

Andrew Pyle