What If I Asked You…?

The Days Ahead:

• Consumer credit and more company reports.

This Week:

• Economic survey response rates are falling…
• We look at why and what it means.
• There’s change coming in the way the University of Michigan measures consumer confidence.
• The data may be very skewed in the next few months.
• The Japanese Yen is very weak and starting to cause international problems…
• What would start its strengthening?
• The Fed didn’t cut but they changed the balance sheet run-off…
• And, rightly, scotched the idea of “stagflation”.

---

What If I Asked You…?

Much economic data relies on survey responses. Some, like the Federal Reserve (Fed) Districts’ monthly surveys, or Beige Book, have very high response rates of 71% to 83% or higher. Anything to do with business and governments tends to receive high response rates. The Census Bureau’s monthly survey of retail sales, for example, goes out to 4,800 firms covering 13,000 units. Companies that do not respond are given an extension, which explains why numbers are often revised. But there are enough responses to provide a 90% confidence level in the results.

Response rates for many economic series are changing. The BLS uses multiple sources and methods from companies and households for its inflation report. In 2003, the BLS sent out 1,117,186 requests to ask about prices for food, clothing, transport and most good and services. It received an 84% response for all nonhousing items.

For housing, the largest contributor to the CPI index, the response rate was 76% and the non-response rate for non-vacant houses (or “Hello, yes I’m here but I’m not going to answer your questions”) was 15%.

In 2023 the response fell to 71% for non-housing items, 65% for housing, and 30% for the non-response rate for non-vacant houses.

People are just not that keen to respond to surveys these days.

The BLS made up for some of this by increasing the number of requests by 9% to 1,215,000 and the number of households by 19% to 81,000. That helps reduce the margin of error.

But surveys from households and individuals are tricky. The chart below shows the response rates for the employment household survey, which is used for the unemployment and participation rates, as well as for job openings.

For the household survey, responses have fallen from a pre-Covid-19 high of 88% to 70% and for the job openings it’s dropped from 66% to 33%.

Both surveys are by phone or in person and for four consecutive months. The responding households are not surveyed for another eight months. After that, they’re not contacted again. This keeps the sample refreshed. The survey is either in person (42%), by phone (49%) or by computer (9%) and the basic questions haven’t changed since 1994.

One of the problems is that the BLS mainly uses phone numbers tied to residences. As we all know, mobile phones are much harder to track down, because people keep their numbers when they move, and we ignore caller i/ds we don’t recognize. An area code may have no connection to where the phone numbers’ owner resides. The BLS is aware of this and uses follow-up e-mails and texts when it can. We’d expect more use of phone apps, messaging and even video data collection.

We’re about to see one significant methodology change for the University of Michigan’s monthly consumer confidence survey. Since 1978, the Fed has used the survey to measure inflation expectations for the next 12 months and the next five to 10 years. The concern is that if consumers expect big price jumps, they will be nervous about the economy and demand wage increases. If they’re comfortable with the outlook for price increases, they feel less inclined to seek wage hikes.

Since 2022, respondents’ expectations of price increases for the next year ranged from 5.5% in 2022 to 2.9% in February. That 5.5% made the Fed nervous. But the same respondents’ expectations in 2023 and 2024 for inflation five years out are more benign, ranging from 3.0% to 3.9%. Here it is (green line) compared to actual inflation.

Since 1979, average inflation has been 3.6% and average expected inflation has been 3.2%. That’s a good match. But in the high inflation years of 1978 to 1982, expectations hit 6.0% and actual inflation was 8.6%. Once the inflation demon was slain, from 1992 on, inflation expectations averaged 2.9% and actual inflation was 2.6%. Inflation expectations are now remarkably stable, even with the spikes in 2022.

It’s an important measure for the Fed but will now change. It’s currently a land line and cell phone interview but will switch to a web-based interview over the next three months. This could be a problem. A typical question in the survey reads like this:

“By about what percent do you expect prices to go (up/down) on the average, during the next 12 months?”

If people reply with “I hope…” the interviewer corrects them with a “No…what do you expect?” If the respondent isn’t comfortable with percentages, the interviewer follows up with

“How many cents on the dollar do you expect prices to go (up/down) on the average, during the next 12 months?”

And if they hear an answer that is out of range, say 6%, they will follow up with:

“Let me make sure I have that correct. You said that you expect prices to go (up/down) during the next 12 months by 6 percent.  Is that correct?”

In other words, it’s a very guided interview. The web-based interview can’t possibly keep up because they can’t hear the intonations or implied questions. The University of Michigan has run the two methods side by side for a while and finding significant differences in the inflation expectations responses, with the web-based responses quite a bit higher and more volatile than the phone responses. In some questions, the differences are as high as 50%. So, as the survey Director says:

“We recommend that data users take a more conservative approach to interpreting change estimates during this transition period.”

No kidding.

We’ll wait and see. The Fed will know about these changes of, course, and build it into their comments. We’d expect some surprise results, in both directions, until the data settles down. Still, with everyone’s eyes on inflation, it’s not coming at the best of times.

Why Is the Japanese Yen So Weak?

The quick answer is because the Bank of Japan (BOJ), who sets rates, and the Ministry of Finance (MOF), who decides if the currency needs support, want it that way.

The scale of the yen’s weakness is hard to overstate. Here’s a quick graph (an up line means a weak yen).

Back in the early 1980s Japan exports were on a tear. Goods exports doubled from 1977 to 1980 and doubled again from 1980 to 1986. As a percentage of GDP they went from 10% to 15%. The yen strengthened from ¥300 to ¥240 in the four years to 1980, which should have slowed export growth, but it stalled out at that level until 1985.

Part of that was dollar strength. After all the U.S. fed funds rate was at 19% in 1981 and even in mid-1984 was 11%. But by 1985, the U.S. authorities had had enough and asked German, French, Japanese and British (it was when sterling mattered, what can I say?) officials over to the New York Plaza Hotel and asked for:

“….some further orderly appreciation of the non-dollar currencies is desirable”

The various central banks then started to sell dollars against their own currency, with the Japanese authorities first out of the gate. The yen appreciated by over 89% in the next three years. The Japanese stock, real estate and everything bubble took another few years to break but by 1990, the Japanese economy had peaked.

For the next 30 years, Japan’s nominal GDP rose from ¥549 trillion to ¥554 trillion or 0.9%. Not per year. Over the entire period. The reason citizens didn’t start storming the castle is because prices dropped and real GDP rose 24% or 0.7% a year. And because the Japanese population declined, GDP per capita grew by around 25%.

The yen strengthened again from 2008 to 2012 as it was seen as a safe-haven, low inflation currency. But for the Japanese authorities it was far from welcome. Growth was zero, rates were zero and deflation was causing major growth problems. The yen peaked at ¥80 in 2012 and gradually fell to ¥106 by 2020. But when rates everywhere except Japan started to rise in 2022, the yen went into free fall. Today it’s at ¥160, about the same as it was at the time of the Plaza Accord.

You would think this would increase Japanese import prices. It has. By about 5% in the last four years. But the BOJ is determined to get inflation above zero, not just for a month or two, but for a prolonged time. The BOJ’s 2% inflation target is the same as the Fed’s but it’s coming at it from the other direction. It wants to break the deflation cycle and see inflation settle at 2%. Its recent peak was 4.3% in 2023 but it’s already down to 1.8%.

The yen has weakened by 11% since January and 15% in the last year against the dollar. That makes Japanese exports very competitive against U.S. dollar-priced products. It also creates problems for the Chinese because the renminbi is mostly pegged to the U.S. dollar. The renminbi is at a 31-year high against the yen at CN¥21, or 15% above a year ago and up 40% since 2020.

How weak is the yen? The BOJ keeps daily track of various real and nominal values of the yen and it’s currently 40% below its purchasing power of the last 55 years. Put more simply, when Japan sells a million dollars of cars, it can only buy 40% as much oil, wheat, gold, or iPhones as it could from 1970 to 2024.  

The strength of the dollar, especially against the yen, may be showing up in U.S. trade. Last week’s GDP number was +1.6% against expectations of +2.5% The 0.9% miss was all in imports, which reduced GDP by 0.96%.

Right now the dollar is way too expensive, whether you use purchasing power parity, real effective exchange rates, relative inflation, or historic prices. But dollar strength may stay that way because U.S. rates are high and likely to remain so. U.S. growth is also the highest among developed countries and every time there’s a geo-political tremor, investors line up to buy dollars.

Japan does not lack the means to intervene in the foreign exchange (FX) markets and revalue the yen.

Japan’s reserves are shown in the green line at $1,246 billion or 30% of GDP. U.S. reserves are at $242 billion or 0.8% of GDP. At any time, Japan’s MOF, could stamp into the FX markets and send the yen 10% higher in a morning. They did it in 2022 when the yen rose (it’s all backward in the FX market, a lower number means a rising yen) from ¥148 to ¥128 in two months.

We’re not market timers, but we think that at some point, Japan will come under intense political pressure, probably from the U.S. and China, to let the yen appreciate. That would be good for U.S. investors in Japan. But for now, it’s a waiting game.   

The Bottom Line

It was a big news week with payrolls, job openings, quit rates, productivity, ISM manufacturing and trade, all A-list economic pointers. Most showed weaker growth but not enough to cause outright concern.

But the Fed’s meeting was the main event. At the start of the year, the markets thought there would be a 99% chance that the Fed would make its first or second cut in this May meeting. But steady growth, solid employment and sticky inflation meant by early April, no one expected a May cut.

And that’s what happened. The Fed left rates unchanged for the sixth successive time since last July. They did, however, change the rate at which the $6,932 billion balance sheet would run off. It peaked at $8,504 billion in May 2022. Later that year, the Fed announced it would run off the Treasury holdings at the rate of $60 billion a month and the Mortgage-Backed Securities (MBS) portfolio at $35 billion a month. That will now change to $25 billion and anything left over from the MBS run off will buy treasuries. The two together mean that there will be around $40 billion less treasuries coming into the market every month.

We shouldn’t make too much of this, however. It’s not really a sign of easier rates ahead but more of an intent to keep the treasury market operating smoothly.

We liked Chair Powell’s burying the idea that the economy is in any form of stagflation, seeing nothing that was either was stagnant or inflationary. Bravo. We’re not remotely close to 1970s conditions.

Friday’s job report came in at 175,000, which was the lowest since October 2023. The unemployment rate barely changed at 3.9%. We expect hiring slowed among smaller companies as the NFIB report has signaled for months about slower hiring.

The Treasury market reacted well, with yields down by about 0.2%. Stocks were fine with the job report and ended up 0.2% for the week and up 3% since the recent lows in mid-April.

Right on cue, the MOF intervened in the foreign exchange markets on Thursday and Friday, taking the yen from ¥158 to ¥155 and then again from ¥155 to ¥152. Clearly the ¥160 level seemed a step too far. We’d expect more intervention.

---

Art of the Week: Louis Finkelstein (1923 – 2000)

---