Couple of weeks ago, we discussed how it doesn’t pay to be “too” smart in the markets. Always compare any new approach to a barebones model to make sure that the extra effort and complexity is worth it.
Whenever a new prediction model gets popular, the first thing people do is to try and use it to predict stock prices. Facebook’s (Meta’s) Prophet library was specifically written for time-series prediction. So, can it be used to construct a momentum portfolio?
Turns out, it could work: Prophet for Momentum. But there’s no urgent reason to use it over a barebones model.
If you take a step back and think about it, aren’t all these momentum models fishing in the same pond? Complexity is then a sales pitch: Returns under Momentum.
Markets this Week
More here: country ETFs, fixed income, currencies and commodities.
Links
Research
The Law and Economics of Mutual Fund Fees (SSRN)
It is well known that fees are the primary cause of the underperformance of actively managed mutual funds. Yet fees have remained stubbornly high.
Powering Past “Peak Passive”? (SSRN)
While increased passive investing is associated with greater market inefficiency, inefficiency may not be an effective mechanism for establishing an active-passive balance. Instead, given the advantages that passive investing may enjoy, the market could well power past “peak passive”.
Equity Premium Events (SSRN)
We use a data-driven approach to identify events that are significantly priced by equity markets without taking a stance on what those events are.
Macroeconomic Expectations and Expected Returns (SSRN)
Using the macroeconomic forecasts of professional economists, we construct a comprehensive macro condition index that summarizes subjective expectations of output, inflation, and labor and housing market conditions. The index predicts stock returns and produces countercyclical equity premium forecasts. We show that the index reflects the true but unobserved macroeconomic condition that impacts the equity premium.
Cross-Country Factor Momentum (SSRN)
Factors in winning countries consistently outperform those in losing countries.
Investing & Economy
What happens if you buy every asset you can get your hands on? Some investors are about to fcuk around and find out: The Cockroach Approach
India
The twin shocks of bad debts in the banking system and corporate de-leveraging followed by the big covid shock caused temporary but significant setbacks to the Indian labour market. But these are on the mend, and recovery in the jobs market has commenced (livemint).
Since December 2019, real GDP has grown by 4.2% at an average annual rate, meaning that India, like many other countries, has not recovered to its pre-pandemic trend. Corporate and foreign investment remain weak. But looked at since December 2021, India’s overall economy seems robust, having grown at 7.1% annually. Alternative indicators, from electricity use to freight traffic, are strong; surveys of purchasing managers for both manufacturing and services have hit their highest levels in over a decade. Forecasters expect 6.5% annual growth over the next five years. (economist)
If India hits all the right economic notes it could lead global growth by 2028 (bloomberg).
70% of 36 popular protein supplements sold in India mislabeled, 14% contain toxins (theprint).
RoW
Someone bet big on a benign US CPI report (bloomberg) and got crushed (bloomberg).
American dysfunction: hyper-concentration in key industries in the name of security, combined with all the perils of short-term financial market pressures which trump (no pun intended) any particular national interest (ft).
Thailand will giving away $14 billion cash in a “digital wallet” program — which will see 50 million adult Thais each getting 10,000 baht ($275). The money is to be spent on goods within a specific time-frame in a designated area (bloomberg).
A fleet of electric-car startups is struggling to stay in business (economist).
A majority of British voters now believe Brexit was a mistake (economist).
A.I.
AI's power appetite could devour 25% of US electricity by 2030 (theregister)
Ironically, using AI might be the best way to find these energy savings. Schneider Electric CEO Peter Herweck predicted that AI could reduce energy consumption in buildings by 15% to 25% over the next four years. (livemint)
After Suno comes Udio - another ChatGPT for Music (rollingstone)
OpenAI, Google and Meta ignored corporate policies, altered their own rules and discussed skirting copyright law as they sought online information to train their newest artificial intelligence systems. (nytimes)