Zombies and New Business Applications

The Days Ahead:

• Earnings season starts with the banks

This Week:

• U.S. entrepreneurialism on the rise
• Soaring new businesses applications
• Term premiums: an opaque concept and not very helpful
• But the Fed is talking about it
• Zombies’ day of reckoning
• Companies that got away with years of zero rates are calling it a day
• Lots of Fed speak this week; mostly “no go” for November hike

Upsurge of New Firms

Over the last few years, there’s been plenty of talk about the fate of the American entrepreneur. One 2014 study worried that not only was the number of new businesses declining but so was the quality. They either didn’t last very long or they were “subsistence” businesses that barely paid one person and had little likelihood of growing. The authors noted that the share of employment by companies has declined by 30% in 10 years.

Why? Well one of the reasons is the changing definition of a “new firm” and the difficulty of keeping track of businesses. You’d think they would be easy to count.

• How long has your company been around?
• Just started six months ago.
• Well done…so we’ll mark you down as a new business.
• Fine. Thanks. Bye.

But what may have happened was that an existing business opened up a new establishment or location. Or the same company spun off its transport division. Or sold its warehouse. Sure, they’re new businesses but they don’t measure the rate of net entrepreneurial activity. They’re just ownership changes.

• It’s us again, sorry to bother. That new company you started…
• Yes
• Did you used to work at Acme Fireworks?
• Yes, we’re very popular in the coyote community.
• Yes, yes, but did you buy their trucks and start a new company and take some mates with you?
• Yes
• Oh, you’re not really a new company, sorry.
• Fine, thanks. Bye. Oh…Roman candle?
• Not today.

Another reason is the difficulty tracking the birth-death model. This is where a company may downsize and a few employees start up a competitor across town and hire a bunch of their old workmates. Yes, it’s a new businesses but it’s not adding net new employment. Most measures of new businesses try to strip out new firms created from old ones.

The 2007 to 2009 financials crisis played havoc with new businesses formations when the annual rate of start-ups went from 12% of all businesses to 8%.

Some studies blamed venture capital (VC) firms, which sounds very counterintuitive.  One study compared the investments of Charles River Investments, a long standing VC company, which in 1970 invested well over half its capital in farming, electric cars (Electro Motion…ahead of its time, here’s one of its cars) and specialized manufacturing and telecom hardware. Today, the Charles River portfolio is nearly all tech, cloud or social network companies. That’s fine but it means the concentration in software and related businesses may have crowded out other start up opportunities.

Overall, VC capital start-ups in the 1980s and 1990s used to put around 45% of portfolio capital in to hardware, material and energy businesses. It’s now less than 5%.  VC companies also tend to be very concentrated with around 5% of all VC firms raising over 50% of all capital in the five years to 2020. The result is that venture funding is effective in stimulating certain kinds of innovative businesses but the scope is increasingly limited.

So new business applications and formation fizzled along at relatively low levels for a number of years. If you were a small company, not really planning to grow and not in tech, starting a new businesses was hard and there wasn’t much financial help. But then in 2020, new business applications took off.

What happened? Covid-19 was a big driver. It allowed people and companies to move out of cities and work from home. It even had a name: the “donut pattern.” Business applications surged in the suburbs as people left city centers, leaving behind long commutes and crowded office space.

The national average rate of new businesses applications grew from 200,000 a month to 550,000 a month, before settling in at around 460,000 today. And these weren’t one-person consulting or shell companies. The proportion labelled as having a “high propensity to grow” or “with planned wages”, meaning they increase hiring and pay people, went from 30% of the total to 42%. Not all applications turn into formations. In the years up to 2019 it was around 9.5% but the latest numbers show that rising to 16%.

And the industries? Construction, transportation and professional services were around 30% of the total. But there was no strong domination by any one industry. It looks like new applications and formations are widespread across professions, locations and pay grades. In other good news, the rate of entrepreneurship among women has risen from 22 per thousand in 2015 to 28 in 2022 and among men from 41 per thousand to 45 per thousand. Immigrants start businesses at twice the rate of native-born people, which is, you know, interesting.

All these patterns are recent and the U.S. Census Bureau and other sources aren’t sure how it will end. But this much was clear. The recent spike in businesses applications is now over three years old. It is potentially,

“A beginning of a reversal of the shift in activity to large and mature firms…and a reversal in pre-pandemic (i.e., poor) business dynamism”

It’s a great number to see and surprising given the chaos of 2020. It also goes some way to explain the recent strong jobs numbers. U.S. entrepreneurialism is back.

Term Premiums

Somewhat reluctantly, we’re going to discuss term premiums. Reluctant because it’s an opaque concept and wheeled out by some commentators and bond specialists on slow news days. Never very convincingly in our view but…people are talking about it.

The 10-year Treasury yield has risen from 0.6% to nearly 4.9% in three years. It also rose very quickly in the last six weeks from 4.1% to 4.9%, although as we write, it’s down to 4.6% following events in Israel. Some of that move is the “Oh the Fed means it” theory, which says, the Fed will likely keep rates higher for longer, given the economy’s resilience with strong job numbers, slowing inflation and no apparent distress in any major businesses sector.

Some of the move is because of the term premium. It’s a tricky measure. It tries to measure how much extra income investors need from a bond for taking on term risk. If the credit risk is constant, as say with a U.S. Treasury, then how much should I expect to receive to take on the risk of a 1 -year Treasury over a 2-year Treasury or over a 3-year Treasury, and so on? In theory, I could just keep buying 1-year Treasuries once a year for 10 years (we’ll stick with the 10-year example now). Or I could buy one 10-year Treasury if it gave me more income or a premium over the one-year Treasury. But how much of a premium?

Enter the term premium model. Normally, you would expect more return on a 10-year bond because buying a series of 1-year Treasury bills 10 times exposes you to all sorts of interest rate risk. Will next year’s rate be higher, in which case it’s a good idea? Or lower? In which case I would have been better off buying the long-term bond. So, I need a premium return for tying up my money for longer,

But if, say, you thought rates were just going to head and stay lower because the economy was really weak, then you may think “I don’t need a premium to hold a bond for 10 years. I’m happy to receive less because I think bond yields are headed lower or even below zero.” In which case the premium goes negative.

How is the number calculated? Therein lies the problem and there are several models. The Federal Reserve model takes the average estimates of what short-term rates will be over 10 years and subtracts that number from the current 10-year yield. You can see a problem right there. Who has a decent track record of predicting rates over 10 years? No one. But we’ll go along with it. So, if estimates are at 5% and the 10-year yield is at 7%, then the term premium is 2%. There is a much better explanation in a 68-page study from the Fed for the pros out there.

The Fed estimate of the term premium is.

The term premium was negative for nearly six years from 2017 but has now risen to a positive number. From 1993 to 2007, the average premium was 1.6% for the 10-year Treasury. If that number on the right, at 0.27%, were to jump to 1.6%, 10-year treasuries would rise to around 6.2%. That would hurt.

There are a number of bond bears who think this is going to happen and give their reasons as high Treasury issuance, U.S. political instability, a drop off in foreign demand for Treasuries and more inflation volatility.

We’d acknowledge that those risks are there but there may be a less complicated reason for the premium going higher. Bond investors haven’t changed their outlook, they just see more of the same: higher rates, a good economy, robust employment and steady growth. In other words, the markets are taking Fed Chair Powell at his word:  

“We’re prepared to raise rates further if appropriate, and we intend to hold policy at a restrictive level until we’re confident that inflation is moving down sustainably toward our objective.”

The term premium today suggests yields at around 4.9%, which is where they recently traded. We don’t know what the term premium is going to do next, and neither do its authors. But the premium measure suggests that Treasury bonds are at least reasonably priced.

As with the “R*” or the natural rate of interest  which we wrote about last week, there are two camps.

One is that Covid-19 and its aftermath haven’t really changed the tradeoff between growth, employment and inflation, and secular stagnation is coming back, in which case bond rates will come down hard.

The other is that it has changed. Strong employment, the fiscal stimulus from the energy investments and re-shoring are all great, and rates will stay higher, and the term premium is a good thing.

We’re in the second camp. We’ll keep a close watch on the term premium data. But try not to take it too seriously for now, it looks fine.

Zombie Invasion

One solid criticism of years of low rates was that unprofitable companies were able to limp along with low borrowing costs. A near zero cost of financing can mask a host of bad management, bad execution, ill-thought-out business plans and “we’ll be profitable one day, promise” ventures. Just ask the crypto guys or WeWork, everyone’s favorite example of a company that was bound to drop to zero the minute its financing costs changed. It managed to fall from a $47 billion valuation to $116m in less than three years.

 If a company has to borrow any money at rates of, say 4%, life is going to be a lot easier to deal with than rates at 10%. So, the worry is that investors may be unwilling to subsidize companies which can’t deliver returns more than their cost of capital.

How many unprofitable companies are out there?

We sorted all the 6,453 companies that showed up in our FactSet data of public companies and found that 1,142 or 17% were unprofitable. But the total sum of their losses was $142 billion and the earnings of the rest of them were $2,868 billion. Of those losses, the top 60 loss makers accounted for $54 billion of the $142 billion, or 38%, and some of those names are perfectly viable companies like Rivian, Moderna, Snap, Western Digital and Stanley Black & Decker.

Ok, so they’re not that important in terms of total losses and many of the companies are probably in temporary loss situations. But there are other economic factors that also count like employment and sales. So, we looked at the number of employees in loss making companies and it was 1.7 million out of 45 million employed by all public companies. They accounted for $713 billion in sales out of $23,057 billion for all public companies. That means the loss-making companies represent around 3% of both total sales and employment. That’s not a trivial number by any means and it’s probably a worse number when we include private companies. But it’s not that high.

This is the latest data we have:

This blue line tracks all Chapter 7, 11, 13 and 15 filings per month and shows quite a jump. The green line is just Chapter 13 filings, which is probably the most extreme one for businesses as it usually means assets are sold off. Other chapters may allow a company to continue to trade. The blue bars at the bottom show business debt growth, which ran as high as 25% during Covid-19 but has since eased to around 2%.  

So, it seems there is some clean out of zombie companies going on and especially since April 2023. But we’re not sure it’s enough to derail the economy. We know that companies will fail, or merge, and we’re not surprised to see bankruptcy filings grow given the interest rate increases. But the effect on GDP and employment seems manageable. Goldman Sachs recently put out a note which looked at all public and private companies and thought that if persistently loss-making companies exited, it would create a drag of around 20,000 on the new jobs report. By way of contrast, last month’s report came in at 336,000. We think 20,000 is too high and one reason is because of the first section in this note: new businesses applications and formations are climbing faster than the bankruptcy filings.

No one likes zombie companies but they shouldn’t be too worrisome in a dynamic economy like the U.S.

The Bottom Line

A number of Fed speakers this week who commented on the recent rate rise.

• Dallas Fed President Lorie Logan: “Less need to raise the fed funds rate”
• Fed Vice Chair Philip Jefferson: “ I will be mindful of the additional tightening in train because of our past rate hikes as I consider whether there is a need to tighten policy further in the future.”
• Atlanta Federal President Raphael Bostic: “I actually don’t think we need to increase rates anymore”
• Fed Governor Christopher Waller: “This is why we have taken forceful steps aimed at reducing inflation—and why we will stay on the job to achieve our objective”
• Fed Governor Michelle Bowman: “…the policy rate may need to rise further and stay restrictive for some time to return inflation to the FOMC’s goal”
• Fed Minutes: “there continued to be downside risks to economic activity and upside risks to the unemployment rate.”

Chalk that up as four to stick, one to keep her options open and the 19 at the meeting to watch and probably stick.

We’d put those together and think that the odds of a November, and probably final hike, just dropped. But it doesn’t terribly matter. The market would probably have muted reactions to either a go/no-go decision. Inflation, growth, fiscal policy, geopolitics and Treasury issuance are probably weighing more on the minds of the bond market right now than another rate hike.

The financial response to the attacks against Israel has generally been quiet. The go-to risk assets that normally react to sudden explosions of violence, the Swiss franc, the yen, gold and oil have all retraced their short-term increases. Perhaps the markets are not taking the events seriously enough but financials markets are usually terrible at predicting or responding to geopolitical events so we’ll have to wait to see what unfolds.

The inflation numbers were good this week with the core-consumer Prices at around a 3.6% annual rate and the Producer Prices at 2.2%. The S&P 500 had a bumpy week but is up over the last five days. Earnings season starts soon.

---

Art: Maria Magdalena Campos Pons (b. 1959)