YWR: Ideas Dinner
Disclosure: These are personal views and market commentary only. Not investment recommendations. For that seek professional help!
From: Ron Dallio <ron.dallio@dwater.com >
Date: 19 October 2023 at 19:42:44 BST
To: Erik @ YWR <erik@ywr.world>
Subject: Ideas Dinner
Erik!
Sorry you couldn’t make it to the ideas dinner this year.
I’ll give you the summary, but also wanted to show you the bone-in rib eyes you missed.
The food was great, but idea-wise everyone was bearish.
Paul, says this reminds him of ‘87 when the Fed over tightened and we got the crash. Which he smashed.
Stevie thinks the market is trading heavy and he doesn’t like the charts. He says maybe we get some Q4 seasonal strength, but then 2024 will be tough and we probably give it back.
You know my views on the end of the dollar regime.
Andreas disagrees with your banks call and is short Lloyds.
Oh, I passed on your Autos call (Untouchable #4), but nobody was into it. Yes, super cheap, but nobody wants to own autos with the UAW strikes and going into a recession next year. You know.
As I said the mood was bearish.
Hard to see any upside. Everyone feels like we’ve had a good year. Now we’re going into 2024 with rising interest rates (TLT crash), WW3 breaking out, high valuations and what is going to be a contentious election (Trump arrest). Europe is dead and China is crashing.
Oh, and a commercial real estate blowing up the banks.
Anything I’m forgetting?
Best to sit on cash, gold and oil stocks.
Those were the top ideas.
Hope you make it next year!
Ron
From: Erik @ YWR <erik@ywr.world >
Date: 19 October 2023 at 22:15:30 BST
To: Ron Dallio <ron.dallio@dwater.com>
Subject: Ideas Dinner
Ron!!!!
The steaks look great!! Hope they’re from Montana.
Really appreciate the feedback on everyone’s views.
Wish I could have been there… sounds like I should have been.
Because………. How do I say this?
Everyone sounds super consensus. Sorry.
Have you seen the BofA surveys?
And on the autos. Totally get it.
Nobody wants to own German car companies paying 8% divies when auto inventories are lowest EVER and we’ve had 3 years of below average car sales already. Totally get it.
And don't want to disturb the bearishness, but wanted to flag this small piece of good news on European car sales.
European car sales are +18% YTD (yes + 18%..despite Europe being ‘dead’) through September. Deutsche is revising UP its 2023 and 2024 car sale estimates…. up...
Basically, same set up as US. We’re coming from super depressed levels, cars are old and there’s lots of catch up potential, but nobody thinks it’s possible…. you know… because of the recession.
Oh and on the US autos. I’m really focused on the Germans, but because of your email I looked at estimates for Ford and GM. Interesting to note that estimates are rising even with the UAW strikes. Probably nothing.
Then back to the market outlook. Rates rising..TLT crashing... Markets are expensive. And Stevie’s view that we had a good year already and give it back next year.
Well.. yes the US is a bit expensive, not crazy though. But then everywhere else is dirt cheap (Europe, UK, China, EM).
Interest Rates? Yes, higher than they’ve been but again, not nosebleed. And earnings estimates are holding in there (rising slightly) and we are looking at 10% growth next year.
I don’t know… Maybe it will be ok.
But who knows and talk is cheap.
YWR Machine Learning
So, I decided to do a statistical analysis of 62 years of S&P 500 data (1960-2022) using the scikit machine learning library for Python.
Historically, do these yields crash the market? Do any of these things really matter?
I created a multi-factor regression to predict the % change in the market over the next 12 months using the following 5 variables:
EPS % Change over next 12 months: How much do earnings grow over the next 12 months? Rational being that markets follow earnings growth. For the backtest I use historical change in earnings. For the predictions I use consensus estimates.
Trailing P/E: It seems like it should matter the valuation of the market at the start of the year. But does it?
Last 12 months price change: Does it matter what the market did the last 12 months? If there has been a big rise, does it give it back in the next year? Or, if there is a big sell off, does it recover the year after?
Local 10 year yield: Does it matter the level of the 10 year interest rates? Maybe when rates are low markets are more likely to go up. And maybe the opposite if rates are high. But is that really what happens?
Change in log of 10 year yield: Maybe it isn’t the level of interest rates but the change over the last 12 months. If rates have been falling over the last 12 months maybe it is bullish. And maybe like now, if 10 years have been rising it is bad. Or is it? I used the log of the yield, because the % change in yields can be multiple hundreds of % coming off zero and it can throw off a regression model. The log smoothes things out a bit.
I fed 60 years of historical S&P 500 data into the computer to build a multi-factor linear regression model.
Then I fed in the current #’s for the S&P (11% EPS growth expected, 19.5x P/E, 16% price move over last year, 4.9% 10 year, 8% move in log of 10 yr).
Predicting a 12M % change is a finnicky thing to estimate (most % changes are small) so it’s not a great model (R squared only 12%).
But do you know what I got?
YWR Machine Learning YE 2024 Price Forecast: $4,462 (+4.3%).
So yes, the S&P’s expensive and rates have increased a bit over the year, but if we get the 11% EPS growth, the market probably grinds higher. That’s what the computer thinks. But who knows?
It’s what I was saying in ‘I see Dead People’. Maybe the S&P chops around for years and goes nowhere. Meanwhile, other markets nobody own outperform and everyone misses out while they stare at the SPX every day.
And FYI, as far as the computer was concerned EPS growth, P/E and change in yields were statistically significant. Last 12 month price change and current 10 year yields were not significant. The computer think these datapoints are irrelevant.
Mystery market for 2024?
And you know what Ron?
Just for fun I fed the YWR Machine Learning Model data on Germany, the UK, Japan and Hong Kong. I only had 20 years of data, not 60, but was surprised at the results.
Do you know what market it liked for 2024? And by a lot!
It will surprise you.
I pasted in the results below.
See you next year!
Erik