“I saw this video of Elon a long time ago, talking about SpaceX [here]. And all these space pioneers were really bashing SpaceX and Elon too. And he was visibly very hurt by that and said these guys were heroes of mine and I wish they would see how hard we’re trying,” said Sam Altman, under increasing scrutiny for his life’s work. He’s the Founder/CEO of OpenAI, creator of ChatGPT4. “I definitely grew up with Elon as a hero of mine. Despite him being a jerk on Twitter, I’m happy he exists in the world. And I wish he would do more to look at the hard work we’re doing to get this stuff right,” said Sam [here], defending himself, as humanity wanders its way into an increasingly uncertain future.
here], signed by countless luminaries from the technology industry and various other fields. Emad Mostaque, CEO of Stability AI, a leader in the generative AI space inked it. Many less prominent researchers did too. “We call on all AI labs to immediately pause for at least 6 months the training of AI systems more powerful than GPT-4.” Elon Musk was a signatory, and of course, he’s building an AI driving system which is remarkable but has a long way to go. Based on my personal experience, it has achieved rough parity with a drunk sixteen-year-old. Or, in economic policy terms, Tesla’s AI has reached singularity with the government’s team responsible for stress testing regional banks. Naturally, it surpassed the Fed’s inflation forecasting team many months ago. And without reliable inflation estimates, it’s no wonder that our central bank’s interest rate forecasts are such a poor indicator of future policy rates. Financial markets are consistently superior to any group of forecasters, even the mob at the Fed who not only have the economy’s most comprehensive data sets, but who can also directly manipulate markets to make their forecasts come true. Such teams of experts are no match for free markets, which are humanity’s first glimpse of a true AI. The collective wisdom of the crowd is a clear superhuman intelligence. That is not to say markets are infallible. But they are infinitely adaptable, resilient, and evolve to exploit our vulnerabilities, fears, secrets. At scale, such systems, move money from weak hands to the strong. And so long as humans are the key decision makers in our economic and political systems, markets will outsmart the best AI systems that our computer scientists develop. But as in all battles for survival, superiority, strong hands will figure out how to harness AI to strengthen our advantage as we trade, invest.“Powerful AI systems should be developed only once we are confident that their effects will be positive, and their risks will be manageable,” read the open letter [
Marcel Kasumovich published a great piece this week on financial crises, the difficulty that the Fed finds itself in, and the path toward narrow banking [here].
Putin says Russia to station tactical nuclear weapons in Belarus, CFTC sues Binance/CZ over regulatory violations, First Citizens acquires all deposits and loans of SVB, Netanyahu halts the judiciary reform process for 1m following wknd protests post the abrupt firing of the defense minister who came out against the reforms, ECB’s Schnabel pushed for ECB statement to say more hiking is possible, Japan PPI Services 1.8% (1.7%e), Germany IFO exp 91.2 (88.3e), US Dallas Fed mfg activity -15.7 (-10e), S&P +0.2%; Fed’s Bullard says bank stress can be addressed by regulatory policy – leaving Fed to target inflation via more hikes, Amazon reportedly considering to buy AMC, Brazil CB mins had hawkish tilt / reiterating hiking cycle can resume, BOJ’s Kuroda confirmed long term infl target is still not met, US C/S home prices 2.55% (2.6%e), US cons conf 104.2 (101e), US Richmond Fed -5 (-10e), S&P -0.2%; Alibaba to split into 6 units, Taiwan president Tsai left for meetings with senior officials in the US, UBS brings back Ermotti as CEO, US says CS is still aiding tax evasion, FDIC considering squeezing biggest banks to plug $23b hole, BOJ’s Uchida suggested YCC adj would not be communicated ahead of time, Biden / McCarthy publicly resume exchanging jabs over the debt ceiling, ECB’s Lane says further rate rises are needed if recent financial stress stays contained, Australia CPI 6.8% (7.2%e), Germany cons conf -29.5 (-30e), Sweden ret sales -9.4% (-7%e), US pending home sales 0.8% MoM (-3%e), Russia IP -1.7% (-1.5%e) / ret sales -7.8% (-8%e), S&P +1.4%; S. Africa CB hikes 50bp (only 25bp exp) / 2023 CPI forecast to 6% from 5.4%, Mexico CB hikes 25bp as exp, Trump indicted on hush payments made to Stormy Daniels / arraignment expected on Tuesday, WSJ reporter detained in Russia on espionage charges, DW/BTFP facility usage drops ~$10b, Finland’s NATO bid clears last hurdle with Turkey’s ratification, Brazil’s fiscal announcements considered ‘less bad’ than feared, Taiwan president lands in NY despite China’s threats, Bolsonaro returns to Brazil to lead the opposition party, EU’s von der Leyen warns of being too tough with China, Fed’s Collins says more hikes needed / Kashkari says services economy is too strong / Barkin uncertain on size of next move, S. Africa Private sector credit 8.28% (8.1%e), Switzerland leading indicator 98.2 (100.8e), Spain CPI 3.1% (3.7%e) / Core CPI 7.5% (7.6%e), Italy unemp 8% (7.9%e), Germany CPI 7.4% (7.3%e), US Jobless claims 198k (196k exp), US 4Q GDP (third) 2.6% (2.7%e), S&P +0.6%; China mfg PMI 51.9 (51.6e) / serv 58.2 (55e) / comp 57 (56.4p), BOJ widens QE target for next quarter to 100-500b yen (from 200-400b), Japan to impose new restrictions on chipmaking gear, ECB’s Villeroy says hikes possibly still have a little ways to go, Fed’s RRP facility usage jumps to 2023 high - $2.375t, Biden pressing for more regulation of midsized banks, Nasdaq finishes Q1 with the best quarter since 2001(!) +19.5%, S. Korea IP -8.1% (-7.5%e), Japan jobless rate 2.6% (2.4%e), Tokyo CPI 3.3% (3.2%e) / Core CPI 3.4% (3.2%e), Japan ret sales 6.6% (5.8%e), UK house prices -3.1% (-2.2%e), Germany ret sales -7% (-5.1%e), Germany unemp 5.6% (5.5%e), EU unemp 6.6% as exp, Italy CPI 8.2% (8.8%e), EU CPI 6.9% (7.1%e) / Core CPI 5.7% as exp, Mexico unemp 2.72% (3.04%e), Brazil unemp 8.6% (8.7%e), US PCE 5% (5.1%e) / Core PCE 4.6% (4.7%e), US Chicago PMI 43.8 (43e), US UofM sentiment 62 (63.3e) / 1y infl exp 3.6% (3.8%e) / 5-10y infl exp 2.9% (2.8%e), S&P +1.5%.
S&P 500 +3.5% and VIX -3.04 at +18.70. Nikkei +2.4%, Shanghai +0.2%, Euro Stoxx +4.0%, Bovespa +3.1%, MSCI World +3.7%, and MSCI Emerging +1.9%. USD rose +1.6% vs Yen, +0.9% vs Russia, +0.6% vs Turkey, and +0.1% vs China. USD fell -4.0% vs Ethereum, -3.5% vs Brazil, -2.2% vs Mexico, -2.0% vs South Africa, -1.8% vs Chile, -1.7% vs Bitcoin, -1.7% vs Canada, -1.1% vs Indonesia, -0.8% vs Sterling, -0.7% vs Euro, -0.6% vs Australia, -0.4% vs India, and -0.2% vs Sweden. Gold -0.8%, Silver +3.5%, Oil +9.3%, Copper +0.5%, Iron Ore +6.5%, Corn +2.7%. 10yr Breakevens (EU +9bps at 2.36%, US +11bps at 2.32%, JP flat at 0.62%, and UK +28bps at 3.78%). 2yr Notes +26bps at 4.03% and 10yr Notes +9bps at 3.47%.
S&P 500 +3.5% and VIX -2.00 at +18.70. Nikkei +2.2%, Shanghai -0.2%, Euro Stoxx -0.7%, Bovespa -2.9%, MSCI World +2.8%, and MSCI Emerging +2.7%. USD rose +3.3% vs Russia, +1.6% vs Turkey, and +0.7% vs Australia. USD fell -17.5% vs Bitcoin, -10.7% vs Ethereum, -3.8% vs Chile, -3.3% vs Brazil, -3.1% vs South Africa, -2.6% vs Sterling, -2.4% vs Yen, -2.4% vs Euro, -1.7% vs Indonesia, -1.4% vs Mexico, -1.0% vs Canada, -0.9% vs China, -0.6% vs Sweden, and -0.6% vs India. Gold +7.2%, Silver +14.6%, Oil -2.0%, Copper +0.1%, Iron Ore +4.5%, Corn +4.8%. 5y5y inflation swaps (EU -15bps at 2.36%, US -6bps at 2.32%, JP -8bps at 0.62%, and UK +18bps at 3.78%). 2yr Notes -79bps at 4.03% and 10yr Notes -45bps at 3.47%.
S&P 500 +7.0% and VIX -2.97 at +18.70. Nikkei +7.5%, Shanghai +5.9%, Euro Stoxx +7.8%, Bovespa -7.2%, MSCI World +7.3%, and MSCI Emerging +3.5%. USD rose +8.3% vs Russia, +4.5% vs South Africa, +2.5% vs Turkey, +1.9% vs Australia, and +1.3% vs Yen. USD fell -42.1% vs Bitcoin, -35.1% vs Ethereum, -7.5% vs Mexico, -6.5% vs Chile, -4.1% vs Brazil, -3.7% vs Indonesia, -2.1% vs Sterling, -1.2% vs Euro, -0.7% vs India, -0.4% vs China, -0.3% vs Canada, and -0.2% vs Sweden. Gold +6.9%, Silver -0.2%, Oil -5.9%, Copper +7.3%, Iron Ore +11.8%, Corn -2.6%. 10yr Breakevens (EU +9bps at 2.36%, US +2bps at 2.32%, JP -23bps at 0.62%, and UK +16bps at 3.78%). 2yr Notes -40bps at 4.03% and 10yr Notes -41bps at 3.47%.
Mexico +20.3% priced in US dollars (+11.2% priced in pesps), Ireland +19.1% priced in US dollars (+17.3% priced in euros), Czech Republic +17.4% in dollars (+12.5% in koruna), NASDAQ +16.8%, Italy +16.1% (+14.4%), Euro Stoxx 50 +15.4% (+13.7%), Greece +15.1% (+13.4%), France +14.8% (+13.1%), Germany +13.9% (+12.2%), Spain +13.8% (+12.2%), Taiwan +13.1% (+12.2%), Netherlands +11.4% (+9.7%), Denmark +11.2% (+9.7%), Sweden +9.5% (+8.8%), Chile +7.7% (+1.2%), Russia +7.3% (+13.8%), Korea +7.3% (+10.8%), MSCI World +7.3% in dollars, S&P 500 +7%, China +6.4% (+5.9%), Japan +6% (+7.5%), Portugal +5.2% (+3.7%), UK +4.5% (+2.4%), Switzerland +4.4% (+3.5%), Austria +4.2% (+2.7%), Belgium +4% (+2.5%), Canada +3.7% (+3.7%), Poland +3.5% (+2%), Argentina +3.1% (+21.6%), Hungary +2.9% (-3.4%), Indonesia +2.9% (-0.7%), HK +2.5% (+3.1%), Russell +2.3%, New Zealand +2.1% (+3.6%), Philippines +1.5% (-1%), Saudi Arabia +1.2% (+1.1%), South Africa +1.2% (+5.3%), Singapore +0.9% (+0.2%), Finland +0.7% (-0.7%), Australia +0.2% (+2%), Thailand -2% (-3.6%), Brazil -3.1% (-7.2%), India -3.4% (-4.1%), Venezuela -4.1% (+38.6%), Malaysia -5% (-4.9%), Norway -6% (+0.1%), Colombia -6% (-10%), Israel -7% (-4.8%), UAE -7.6% (-7.6%), and Turkey -14.8% (-12.6%).
A human player defeated a top-ranked AI system at the board game Go, in a surprise reversal of the 2016 computer victory that was seen as a milestone in the rise of artificial intelligence. Kellin Pelrine, an American player who is one level below the top amateur ranking, beat the machine by taking advantage of a previously unknown flaw that had been identified by another computer. But the head-to-head confrontation in which Pelrine won 14 of 15 games was undertaken without direct computer support. The strong hand won, and won, and won.
The tactics were suggested by a computer program that probed the AI systems looking for weaknesses. The software played more than 1mm games against KataGo, one of the top Go-playing systems, to find a blind spot” theat a human player could take advantage of. The suggested plan was then ruthlessly executed by Pelrine. The tactic is very rare, indicating the AI systems had not been trained on enough similar games to tackle the situation. The same tactic would not have beaten a human as they would have adapted to an unfamiliar circumstance.
In Go, two players alternately place black and white stones on a board marked out with a 19x19 grid, seeking to encircle their opponent’s stones and enclose the largest amount of space. The huge number of combinations means it is impossible for a computer to assess all potential future moves. The tactics used by Pelrine involved slowly stringing together a large loop of stones to encircle one of his opponent’s own groups, while distracting the AI with moves in other corners of the board. A human would have found it quite easy to spot.
“The discovery of a weakness in some of the most advanced Go-playing machines points to a fundamental flaw in the deep-learning systems that underpin today’s most advanced AI,” said Stuart Russell, a computer science professor at the University of California, Berkeley. “The systems can understand only specific situations they have been exposed to in the past and are unable to generalize in a way that humans find easy,” he added. “It shows once again we’ve been far too hasty to ascribe superhuman levels of intelligence to machines.”
“Before globalization and the internet, everyone knew the world’s top three best tenors,” said Lindsay Politi, our inflation specialist. I naturally knew none. “The distinction between the second-best tenor and fifth best was relatively small,” she added. “Today there are huge economic windfalls to being the first, second and third best tenors in the world and almost none to being the fourth best. A huge chasm between the top few and everyone else opened. This chasm in many ways has been one of the defining features of the information age.”
“This effect has already almost fully played out in manufacturing,” continued Lindsay. “AI brings the fourth tenor effect to white-collar jobs. The economic value of mediocre and derivative work is plummeting. But the value of truly excellent and original work will become even more unique, more highly prized. A chasm between the third best and fourth best script writer, graphic designer, lawyer, etc. is just starting to open. ChatGPT might write the next Marvel movie but will never write the next Everything Everywhere All at Once.”
“In ten years, when we look back at this remarkable period, what will we wish we had started building today?” I asked our investment team. We carve out time each week with no structure, to see what fills the space. One River had just celebrated its 10-year anniversary with seven volatility-trading and trend-following strategies, alpha strategies, bespoke diversifying solutions, and over $3bln in institutional assets. I reminded the team that in 2013, having spent 24yrs as a discretionary global macro trader/investor, I concluded that if we failed to build a strong quantitative capability at the firm, we’d be unlikely to survive a decade. We looked for opportunities to codify trading/investing and risk-management principles developed in our discretionary activities and fused them into quantitative strategies. They behaved differently from products that we observed at purely quantitative firms. That had been our hope and expectation. And the process led to the development of powerful screening tools that in turn helped our discretionary trading activities. The interaction between discretionary and quantitative investing reinforced a culture of collaboration and exploration that continues as we hunt for behaviors in markets that can be codified and think about risks that reside outside of 20-year data sets. That means nothing stands still, but instead evolves, which is what every investment firm must do to survive, thrive. “The advances in AI feel overhyped right now, which is natural in a tech cycle, with the media amplifying everything these days. I wouldn’t be paying insane prices for AI companies now. But like crypto and blockchain, AI is here to stay, and this will profoundly impact the world.” As with our quant investing build-out a decade ago, and our headfirst dive into crypto in 2020, we need to run toward change. Keep your mind wide open. Eyes and ears too. The markets will give us hints of where this is going. The strong hands will harness these powers, melding man and machine in unprecedented ways. “Our job as investors is to figure out what and how to build for the coming decade and make a ton of money from it.”
Good luck out there,
Chief Investment Officer
One River Asset Management