ON DECK FOR TUESDAY, SEPTEMBER 15th

KEY POINTS:

  • Stocks and bonds continue to waver amid higher oil prices, pre-FOMC
  • Canada’s Investment Summit kicks into high gear as a downpayment on $1 trillion
  • Scotia’s annual Fixed Income Forum in Toronto is today…
  • ...with former PM Stephen, former Ambassador Hillman in the line-up
  • China’s struggling economy…
  • ...and why it wants to invest in places like Canada
  • UK labour market remains soft
  • Canadian existing home sales were little changed
  • Light US data: Empire, weekly ADP
  • AI researchers warn that the end is nigh…
  • ...so I asked ChatGPT for a list of spectacular tech forecast failures

Markets have become the proverbial broken record that spins on the same track over and over. Oil is up again but not by much as Brent sits at US$106 and WTI is at $102. So are longer-run bond yields on inflation fears. Juicy market pricing for central bank rate hikes positions each of the Fed, BoC, ECB, BoE and BoJ either fully or significantly priced for 50bps of hikes from now until year end. Stocks are falling basically everywhere but not by much and on rate and inflation fears plus AI fears either having to do with concern about slowing momentum or the cheery news from AI researchers that all profits will go to zero by decade’s end as the least of humanity’s fate.

On the latter point, you could turn AI onto the AI forecasters themselves by saying to ChatGPT “Please tell me historical examples of tech sector themes, theories and forecasts that have been awfully wrong.” I’m disinclined to go with Trump’s word that recent fears are just scaremongering because, well, it’s Trump, but it is a sector that is highly prone to making grossly exaggerated claims that often flop. I’ve pasted the list at the back of this note. Enjoy, while I look up bunker plans…

Overnight developments were fairly light and recapped below (China and UK data, Canadian home sales). It’s show time for Canada’s Investment Summit today which will mark progress toward unleashing a wave of investment in major projects. I view the list of projects that are a combination of being already in the works, reasonably expected and hypothetical wishes to be a downpayment on PM Carney’s C$1 trillion investment goal over the next five years. Canada saw foreign direct investment rise by about $1½ trillion over the past five years. A Fall Budget is needed to inform steps on financing, including teasers about tax reforms.

I’m looking forward to Scotia’s annual Fixed Income Forum in Toronto which brings numerous speakers and panelists such as former PM Stephen Harper and former Canadian ambassador to the US Kristen Hillman. I’ll give an economic outlook for my part.

CANADIAN HOME SALES POST FIRST DROP IN FIVE MONTHS

Canadian existing home sales posted their first dip in five months during August (-0.7% m/m SA). They climbed by 0.5–0.7% m/m SA for three of the prior four months and a whopping 5.5% m/m gain was registered in May. It's possible that after four straight gains including one whopper that the market had to take a breather. It's also possible that buyers are substituting toward surging new home sales on preferential tax changes that have shaved tens of thousands off the average new home contract price. New listings were up by 3.3% m/m, the sales-to-new-listings ratio fell two full points to 49.1% and remains roughly in balanced territory, and months supply was unchanged at 4.8.

CHINA’S STRUGGLING ECONOMY

China data was fairly soft on balance for the month of August as follow up to the weak financing figures from the prior day.

  • Why is China interested in investing elsewhere including in Canada? Because investment in China itself is in freefall (chart 1). Investment remains weak with total investment down -7.2% ytd/ytd. Property investment is down about 20% ytd/ytd.
  • New and resale home prices continue to slide (chart 2). Potential home buyers won't borrow if it means catching a falling knife via capital losses. 
  • The jobless rate ticked up to 5.3% but is trendless along a curiously oscillating path over the past couple of years.
  • Retail sales have basically stalled (+0.4% y/y). They slipped by -0.1% m/m SA in August, extending a trend of basically no growth for an extended period (chart 3).
  • Industrial output grew faster (5.2% y/y, 4.5% consensus). Chart 4. It was up by 0.5% m/m SA in August and has been a rare bright spot in China’s economy partly due to excess production and exporting this surplus to keep factories busy.

THE UK’S SOFT LABOUR MARKET

The UK labour market remains soft but the net takeaway from soft jobs and resilient wages is unlikely to be a deciding factor for the Bank of England’s policy stance amid expectations for a hold on Thursday with markets priced for 50bps of hikes over the next two meetings.

  • Payroll employees fell again (-26k) in August. It’s the seventh straight decline (chart 5).
  • Total employment lags payrolls by a month and has not been falling, but it has been basically flat for the past two months (+9k in July, -6k in June). Chart 6.
  • Wage growth is volatile but cooled to 3 ½% m/m SAAR in July. The three-month moving average sits at 4.7% m/m SAAR (chart 7).
  • Job vacancies have been holding steady around the long-term average for several months now (chart 8).

AN EMPTY US CALENDAR BEFORE THE FED HIKES TOMORROW

The US calendar is light again on the eve of a widely expected hike by the Federal Reserve. Only the Empire manufacturing gauge of conditions around the NY Fed's district (8:30amET) and the weekly ADP private payrolls measure (8:15amET) are due. 

SPECTACULAR FORECAST FAILURES FROM THE WORLD OF TECH

As noted, I turned GenAI onto itself by asking it to tell me about spectacular claims made by the tech sector over the years that were dead wrong. I’ve only listed the examples without the lengthy colour that was provided. If interested, you can replicate the prompt and get the further colour with elaborations.

ChatGPT prompt: Please tell me historical examples of tech sector themes, theories and forecasts that have been awfully wrong.

ChatGPT’s answer:

There is a surprisingly rich history of technology investment narratives that became consensus themes and then proved badly wrong. The most interesting cases aren’t merely silly predictions; they are cases where the underlying technology was real, but investors and forecasters extrapolated the wrong variable, the wrong timing, or the wrong economic model.

Here are some of the best examples, particularly relevant if you’re thinking about today’s AI/technology investment cycle.

1. The telecom “bandwidth shortage” — late 1990s

2. “The Internet will destroy traditional media” — 1990s/2000s

3. The “paperless office” — 1970s–1990s

4. The PC would not become a mass consumer product

5. Kodak and digital photography

6. 3D television — roughly 2009–2013

7. Netbooks would replace PCs

8. The smartphone would destroy the PC much faster than it did

9. “The cloud will eliminate the need for hardware”

10. Internet traffic forecasts — consistently too high

11. “The metaverse is the next Internet”

12. NFTs / “digital ownership”

13. “SaaS means software margins can only go up”

14. Moore’s Law would continue indefinitely

15. The “AI will replace almost all software engineers immediately” thesis