“The distance between what you forecast and what actually happened.”
Not to be confused with Tim Horton’s Brier Cup (or what Canadian Curling fanatics call The Brier), your Brier Score is a probability weighted measure of how good you are at forecasting.
As Phil Tetlock explained in the must-read #behavioral economics book I’ve been citing for the last month, Superforecasting, “Brier scores are like golf scores: lower is better. Perfection is 0.” (pg 64)
No one is perfect. But that doesn’t mean we shouldn’t strive for excellence. We’ve been within 20-30 basis points of forecasting US GDP perfectly for the last 5 quarters. If we’re right on Q2, being positioned for lower-for-longer is going to make for better performance again.
Back to the Global Macro Grind…
But, but… bond yields rose (for a week), gold is getting “hammered” (to +15% YTD), and… if the SP500 can hold it’s short-squeeze day of +1.4% (yesterday), US stocks are going to be up for the 1st week in the last 6…
Doesn’t that mean GDP is going to be 2.5% and the Fed can raise rates in June?
C’mon. Let’s keep it real here. Inasmuch as the SP500 signaled immediate-term TRADE oversold (after 4 straight down weeks) in the 2030-2040 range last week, it will signal immediate-term TRADE overbought in the 2080-2090 range this week.
Oh, and it’s month-end.
That’s when everyone @CNBC who is in the business of marketing “but the market isn’t down” (meanwhile most fund managers they interview are) gets to try to tell you another story about how #TheCycle wasn’t really happening, afterall.
Yes, getting the rate of change in GROWTH And INFLATION right is a major component of having a low Brier Score. But risk managing the range (i.e. trading) of immediate-term oversold vs. overbought is a critical component to scoring well too.
From a Hedgeye #Process perspective, what I mean by that is this:
- Start with your best Bayesian bet on where the intermediate-term TREND is going (our research view)
- Contextualize that TREND within the long-term TAIL duration (i.e. #TheCycle)
- And then just risk manage the immediate-term TRADE range of oversold/overbought signals
I know. This isn’t perfect. But after 17 years of trial and error (making legions of mistakes), this process of marrying my multi-duration cycle research with quantitative risk management signals (multi-duration, multi-factor) is the best I can do.
What do you do? If you’re reading this, I certainly appreciate you having an open mind to what it is that I do.
What I don’t do is chase high and freak-out (sell) low. While it seems like forever ago (in inbox question terms), it was only last week that I wrote to you about 9-10 buy/cover signals, taking Real-Time Alerts to 9 LONGS and 3 SHORTS.
In the spirit of the chapter in Superforecasting that today’s quote about Brier Scores comes from (called Keeping Score), on May 18th, I also took our “Cash” position down to its lowest level of 2016 at 49%.
That’s when bond yields tapped the top-end of my risk range (and stocks sold off to the low-end of the immediate-term TRADE range) though. This morning European, Japanese, and US Equities are trading at the top-end of the range:
- SP500 immediate-term risk range = 2037-2084
- Nikkei immediate-term risk range = 16,331-16,890
- German DAX immediate-term risk range = 92
So I’ll raise my “Cash” position back up to 58% now (selling some US Equity and USD exposure).
Since Utilities (XLU) and Long-term Bonds (TLT) already signaled immediate-term TRADE oversold (lower) last week, the only big thing that I like on the long side that is signaling oversold today (i.e. it’s at the low-end of the risk range today) is Gold.
Gold’s immediate-term risk range is now $1. So I’d buy some at $1215 and sell some at $1260.
Selling and/or buying “some” in something doesn’t mean I don’t have “conviction.” It means I have a process to measure the distance between what I’m forecasting and what is actually happening.
Our immediate-term Global Macro Risk Ranges (with intermediate-term TREND research views in brackets) are now:
UST 10yr Yield 1.68-1.89% (bearish)
SPX 2037-2084 (bearish)
RUT 1090-1140 (bearish)
NASDAQ 4 (bearish)
Nikkei 160 (bearish)
DAX 92 (bearish)
VIX 13.86-17.58 (bullish)
USD 94.11-95.95 (bullish)
EUR/USD 1.11-1.13 (bearish)
YEN 108.45-110.79 (bullish)
Oil (WTI) 46.98-49.89 (bullish)
Nat Gas 1.95-2.17 (bearish)
Gold 1 (bullish)
Copper 2.02-2.10 (bearish)
Best of luck out there today,
Keith R. McCullough
Chief Executive Officer