Finance

Listen to this story. Enjoy more audio and podcasts on More

Listen to this story. Enjoy more audio and podcasts on More

Listen to this story. Enjoy more audio and podcasts on More

Economists polled by Reuters had expected a decline in the headline consumer price index to 4.4%, after the 4.6% annual reading of October surprised to the downside by dropping to a two-year low.
The Office for National Statistics said the largest downward contributions came from transport, recreation and culture, and food and non-alcoholic beverages.

LONDON, UK – Sept. 2021: People seen dining outdoors in Soho in London in September 2021.
SOPA Images | LightRocket | Getty Images

LONDON — U.K. inflation fell by more than expected in to hit 3.9% in November, in the lowest annual reading since September 2021.
Economists polled by Reuters had expected a modest decline in the headline consumer price index to 4.4%, after the 4.6% annual reading of October surprised to the downside by dropping to a two-year low.

Month-on-month, headline CPI fell by 0.2%, compared to a consensus forecast of a 0.1% increase.
Core CPI — which excludes volatile food, energy, alcohol and tobacco prices — came in at an annual 5.1%, well below a 5.6% forecast.
The surprisingly large falls prompted a spike in bets that the Bank of England will cut interest rates in 2024, which manifested in a sharp fall in British bond yields.
The U.K. 10-year gilt yield sunk to an eight-month low, dropping 11 basis points to around 3.54%. Yields move inversely to prices. Meanwhile, the U.K.’s FTSE 100 was the only major European stock index in positive territory on Wednesday, climbing 0.8% by mid-morning trade.
The Office for National Statistics said the largest downward contributions came from transport, recreation and culture, and food and non-alcoholic beverages.

The Bank of England last week maintained a hawkish tone as it kept its main interest rate unchanged at 5.25%. The Monetary Policy Committee reiterated that policy is “likely to need to be restrictive for an extended period of time.”

The central bank ended a run of 14 straight interest rate hikes in September, as policymakers looked to wrestle inflation back down towards the Bank’s 2% target from a 41-year high of 11.1% in October 2022.
U.K. Finance Minister Jeremy Hunt cheered the Wednesday figures and said the country was “starting to remove inflationary pressures from the economy.”
“Alongside the business tax cuts announced in the Autumn Statement this means we are back on the path to healthy, sustainable growth,” he said in a statement.
“But many families are still struggling with high prices so we will continue to prioritise measures that help with cost of living pressures.”
Significant fall ‘undermines’ Bank of England caution
The Bank of England has repeatedly pushed back against market expectations for significant cuts to interest rates in 2024, noting last week that “key indicators of U.K. inflation persistence remain elevated.”
Suren Thiru, economics director at ICAEW, said the “startling” fall in inflation recorded on Wesdnesday will reassure households that there is a “light at the end of the tunnel,” with easing core CPI figures showing that underlying price pressures are relenting.
“The likely squeeze on wages from rising unemployment and a stagnating economy should help to continue to keep them on a downward trajectory,” he said by email.

“These inflation numbers suggest that the Bank of England is too pessimistic in its rhetoric over when interest rates could start falling. A deteriorating economy could push the Bank to start loosening policy by the Autumn, particularly if inflationary pressures continuing easing.”
A ‘glimmer of relief’
Richard Carter, head of fixed interest research at Quilter Cheviot, said the latest inflation print adds to a sense of “cautious optimism” in the U.K. relative to the cost of living crisis and bond market chaos of last year.
Despite the drop in CPI, he noted that the broader economic picture remains “complex, marred by stagnation and subdued growth prospects.”
The U.K. economy contracted by 0.3% month-on-month in October, after flatlining in the third quarter.
“This stagnation, leaving the output no higher than it was in January, paints a picture of an economy struggling to rebound from a series of unprecedented challenges,” Carter said over email, while acknowledging that the pace at which inflation is slowing offers a “glimmer of relief” for households.
“The pressures are manifold – from the cost of living crisis, volatile energy markets, Brexit aftershocks, to enduring productivity issues. These factors have collectively dampened economic prospects and consumer confidence.” More

Tesla cut prices for its electric cars in China by more than BYD did for its flagship Han sedan, according to analysis from U.S.-based firm JL Warren Capital.
The Han sells in a similar price range as Tesla’s cars — above 200,000 yuan ($28,000).
Most of BYD’s many other cars cost much less.

BYD’s Han electric car, pictured here at the 2021 Shanghai auto show, is one of the most popular new energy vehicles in China.
Evelyn Cheng | CNBC

BEIJING — Tesla cut prices for its electric cars in China by more than BYD did for its flagship Han sedan, according to analysis Wednesday from U.S.-based firm JL Warren Capital.
Tesla reduced the price of its Model 3 by 6% compared to December last year, and cut the price of Model Y by 11% during the same period of time, JL Warren Capital CEO and Head of Research Junheng Li said in the report.

BYD’s Han only saw a 5% price decrease during that time, she said.
The Han, the company’s premium electric sedan, sells in a similar price range as Tesla’s cars — above 200,000 yuan ($28,000). Most of BYD’s other cars cost much less.
The report showed that BYD increased its sales promotions throughout the year, shaving 10% or 17% off the price of some mass market models. “Double-digit discounts are a common promotion by [original equipment manufacturers] to stimulate sell-through and meet the sales target,” Li said.

High-end electric car startup Nio also cut prices this year, despite initially trying to avoid getting caught up in an industry price war.
“Unlike in the EU or the US, residual values do not appear to feature highly in Chinese consumers’ purchase decisions,” HSBC analysts said in a Dec. 4 report about the auto industry. “That is perhaps the reason why price competition is so severe in China relative to EU/US.”

Thanks partly to government support, penetration of new energy vehicles, which include battery and hybrid-powered cars, has surged to well over one-third of new passenger cars sold in China.
Li expects that penetration rate will be around 40% next year, while electric car sales grow by 20%, a slowdown from a 35% increase in 2023.

Read more about electric vehicles, batteries and chips from CNBC Pro

Already for this year, the industry’s largest automakers had an “overly ambitious goal” of 93% sales growth, Li said. She pointed out that among 13 major EV manufacturers in China, only Tesla and Li Auto are set to reach their respective sales targets for the year.
That signals competition is about to get fiercer in China, the world’s largest auto market, which could lead to the potential for industry waste.
“New models spur EV demand, but at the cost of intensifying [the] pricing war as the market is flooded with inventory of ‘obsolete’ models,” Li said, noting the new car development cycle in China has been reduced to one or two years versus about three years previously.  More

Alibaba Group CEO Eddie Wu is taking over the top role at the company’s Taobao and Tmall e-commerce business, replacing Trudy Dai in the Chinese internet tech giant’s latest management shakeup this year.
Wu replaced Daniel Zhang as the group’s CEO in September.
Wu also became acting chairman and CEO of Alibaba’s Cloud Intelligence Group in September after Zhang abruptly left the business unit.

Trader works at the post where Alibaba is traded on the floor of the New York Stock Exchange (NYSE) in New York City, U.S., March 28, 2023. REUTERS/Brendan McDermid
Brendan Mcdermid | Reuters

BEIJING — Alibaba Group CEO Eddie Wu is taking over the top role at the company’s Taobao and Tmall e-commerce business, replacing Trudy Dai in the Chinese internet tech giant’s latest management shakeup this year.
Dai, who is one of the 18 cofounders of Alibaba, will assist in establishing an asset management company, according to an internal letter from Alibaba Chairman Joe Tsai seen by CNBC.

Alibaba’s announcement Wednesday comes after Wu replaced Daniel Zhang as the group’s CEO in September.
Wu has been chairman of Taobao and Tmall Group since May 2023.

The e-commerce business that once propelled Alibaba to success has run into challenges with rising competitors such as PDD, while consumption growth in China remains sluggish.
PDD’s U.S.-listed shares have gained more than 80% so far this year, driving the company’s market capitalization higher than Alibaba’s. In contrast, the company founded by Jack Ma has seen its shares fall by about 14% year to date.
Contributing to a recent decline in Alibaba shares was news last month that the company had scrapped plans to list its cloud business due to U.S. restrictions on exports of advanced chips to China.

Alibaba in March had announced a massive restructuring into six units, paving the way for individual stock listings, especially for its cloud business.
Wu became acting chairman and CEO of Alibaba’s Cloud Intelligence Group in September after Zhang abruptly left the business unit.
“Eddie’s leadership of both Alibaba Cloud and [Taobao and Tmall Group] will ensure total focus on, and significant and sustained investment in, our two core businesses of cloud computing and e-commerce, as well as enabling TTG to transform through technology innovation,” Tsai’s letter said.
“Soon, we will empower a new cohort of management leaders who have developed fundamental skillsets and experience from the bottom up.”
Dai “accomplished” the company’s mission regarding Taobao and Tmall, and her new role in the asset management company would allow her to “play to her strengths,” the letter said.
During Alibaba’s latest earnings call in mid-November, the company said it planned to monetize its non-core assets and noted it had $67 billion on its balance sheet in equity securities and other investments.
Tsai’s letter did not provide details on those non-core assets. More

The number of “accredited” investors swelled to 24 million in 2022, the SEC said. That’s 8 million more than in 2019, and the number is poised to keep growing.
Accredited investors can buy private securities such as private equity, hedge funds and venture capital funds. They generally meet financial requirements tied to net worth or annual income.
Private investments used to be earmarked for roughly the top 2%. Now, about 1 in 5 households can buy them.

Morsa Images | Getty Images

Inflation has given millions of people new access to certain investments earmarked for the wealthy — and consumer advocates argue that’s not a good thing.
Americans must generally be “accredited” to invest in private companies and investments such as private equity and hedge funds.

That accredited status is a consumer protection issue: To qualify, households must meet certain requirements — like a minimum net worth or annual income — which helps ensure they’re financially sophisticated and can sustain the risk of loss from private investments.
Over 24 million U.S. households — about 18.5% of them — qualified as accredited investors in 2022, the Securities and Exchange Commission said in a report issued Friday.
That’s an increase of about 8 million households from 2019, the last year for which the SEC published an estimate. That year, 13% of households qualified.
The increase is “largely due to” inflation, the SEC said.

How inflation affects accredited investor ranks

Individuals can generally become accredited by having a $200,000 annual earned income, or $300,000 for married couples. Individuals or couples can also qualify with a total $1 million net worth, not including the value of their primary residence.

However, those financial thresholds aren’t pegged to inflation. They stay the same even as wealth and incomes naturally grow over time — meaning more people have gradually become accredited over the years.
Indeed, the thresholds haven’t changed since their creation in the early 1980s. In 1983, just 1.5 million households — 1.8% — qualified as accredited investors, according to SEC data.

Most Americans will join the ranks of accredited investors in coming decades if the financial thresholds remain unmoored from inflation: By 2052, nearly 119 million households would qualify — or about 66% of them, the SEC said.
“The pool keeps increasing,” said Micah Hauptman, director of investor protection at the Consumer Federation of America, a consumer advocacy group. “If we don’t do anything, the standard will be rendered meaningless.”
If the financial standards had been indexed to inflation since the 1980s, a married household would need a roughly $3 million net worth or a $911,352 joint income to be accredited in 2022, the SEC said. Just 5.7% of households — about 7.4 million — would qualify, according to its data.
More from Personal Finance:Even high earners consider themselves ‘not rich yet,’ despite their net worthThe S&P 500 is up about 23% year to date. Here’s what to knowOnly 60% of student loan borrowers made payments when bills restarted

The difference between public and private investments

Private investments differ from their publicly available counterparts.
Public investments include ones with which most households are familiar, such as the stocks and funds available for purchase on a stock exchange. Generally, anyone can buy them.
Private investments let people invest in companies that aren’t listed on a public exchange.
Some argue that private investments should be available to a broader pool of investors due to benefits such as higher average returns.
Private equity returns, for example, have outperformed the S&P 500 stock index by 1% to 5% on an annualized basis since 2009, according to a 2021 report by Michael Cembalest, chair of market and investment strategy for J.P. Morgan Asset & Wealth Management.

Others argue that private markets are less transparent, with information about companies and funds less readily available to many investors, and carry additional risks.
“Without information, you have no ability to value the company to make an informed investment decision,” Hauptman said. “You’re investing blind.”
Private investments are also generally illiquid, and investors should be prepared to lock up their money for maybe 10 years in some cases, said Paul Auslander, a certified financial planner and director of financial planning at ProVise Management Group in Clearwater, Florida. That longer holding period could make them riskier for some investors, he said.
“It’s like any other investment,” Auslander said. “You have to read the fine print and make sure you know what you’re investing in.”

Shift away from pensions helps investors qualify

Aside from inflation, trends like the move toward 401(k) plans and away from pensions have contributed to the swelling ranks of accredited investors over time, according to the SEC.
About 85 million people actively participated in 401(k)-type plans in 2020, about three times the number in 1982, the SEC said. Such private retirement savings is included in calculations of net worth.

The pool keeps increasing. If we don’t do anything, the standard will be rendered meaningless.

Micah Hauptman
director of investor protection at the Consumer Federation of America

The shift from pensions may have also “created investor protection considerations” that weren’t present in the early 1980s, according to the SEC. That’s because the responsibility for investment decision-making shifts from employers to individuals, who may lack the experience to appropriately manage investment risk, the SEC said.
There would be about 5 million fewer accredited investors in 2022 if retirement savings were omitted from the net-worth calculation, the SEC said. More

Large language models, similar to the one at the heart of ChatGPT, frequently fail to answer questions derived from Securities and Exchange Commission filings, new research finds.
The findings highlight some of the challenges facing AI models as big companies, especially in regulated industries like finance, seek to incorporate cutting-edge technology into their operations, whether for customer service or research.
“That type of performance rate is just absolutely unacceptable,” Patronus AI cofounder Anand Kannappan said. “It has to be much much higher for it to really work in an automated and production-ready way.”

Patronus AI cofounders Anand Kannappan and Rebecca Qian
Patronus AI

Large language models, similar to the one at the heart of ChatGPT, frequently fail to answer questions derived from Securities and Exchange Commission filings, researchers from a startup called Patronus AI found.
Even the best-performing AI model configuration they tested, OpenAI’s GPT-4-Turbo, when armed with the ability to read nearly an entire filing alongside the question, only got 79% of answers right on Patronus AI’s new test, the company’s founders told CNBC.

Oftentimes, the so-called large language models would refuse to answer, or would “hallucinate” figures and facts that weren’t in the SEC filings.
“That type of performance rate is just absolutely unacceptable,” Patronus AI cofounder Anand Kannappan said. “It has to be much much higher for it to really work in an automated and production-ready way.”
The findings highlight some of the challenges facing AI models as big companies, especially in regulated industries like finance, seek to incorporate cutting-edge technology into their operations, whether for customer service or research.
The ability to extract important numbers quickly and perform analysis on financial narratives has been seen as one of the most promising applications for chatbots since ChatGPT was released late last year. SEC filings are filled with important data, and if a bot could accurately summarize them or quickly answer questions about what’s in them, it could give the user a leg up in the competitive financial industry.
In the past year, Bloomberg LP developed its own AI model for financial data, business school professors researched whether ChatGPT can parse financial headlines, and JPMorgan is working on an AI-powered automated investing tool, CNBC previously reported. Generative AI could boost the banking industry by trillions of dollars per year, a recent McKinsey forecast said.

But GPT’s entry into the industry hasn’t been smooth. When Microsoft first launched its Bing Chat using OpenAI’s GPT, one of its primary examples was using the chatbot quickly summarize an earnings press release. Observers quickly realized that the numbers in Microsoft’s example were off, and some numbers were entirely made up.

‘Vibe checks’

Part of the challenge when incorporating LLMs into actual products, say the Patronus AI cofounders, is that LLMs are non-deterministic — they’re not guaranteed to produce the same output every time for the same input. That means that companies will need to do more rigorous testing to make sure they’re operating correctly, not going off-topic, and providing reliable results.
The founders met at Facebook parent-company Meta, where they worked on AI problems related to understanding how models come up with their answers and making them more “responsible.” They founded Patronus AI, which has received seed funding from Lightspeed Venture Partners, to automate LLM testing with software, so companies can feel comfortable that their AI bots won’t surprise customers or workers with off-topic or wrong answers.
“Right now evaluation is largely manual. It feels like just testing by inspection,” Patronus AI cofounder Rebecca Qian said. “One company told us it was ‘vibe checks.'”
Patronus AI worked to write a set of over 10,000 questions and answers drawn from SEC filings from major publicly traded companies, which it calls FinanceBench. The dataset includes the correct answers, and also where exactly in any given filing to find them. Not all of the answers can be pulled directly from the text, and some questions require light math or reasoning.
Qian and Kannappan say it’s a test that gives a “minimum performance standard” for language AI in the financial sector.
Here’s some examples of questions in the dataset, provided by Patronus AI:

Has CVS Health paid dividends to common shareholders in Q2 of FY2022?
Did AMD report customer concentration in FY22?
What is Coca Cola’s FY2021 COGS % margin? Calculate what was asked by utilizing the line items clearly shown in the income statement.

How the AI models did on the test

Patronus AI tested four language models: OpenAI’s GPT-4 and GPT-4-Turbo, Anthropic’s Claude2, and Meta’s Llama 2, using a subset of 150 of the questions it had produced.
It also tested different configurations and prompts, such as one setting where the OpenAI models were given the exact relevant source text in the question, which it called “Oracle” mode. In other tests, the models were told where the underlying SEC documents would be stored, or given “long context,” which meant including nearly an entire SEC filing alongside the question in the prompt.
GPT-4-Turbo failed at the startup’s “closed book” test, where it wasn’t given access to any SEC source document. It failed to answer 88% of the 150 questions it was asked, and only produced a correct answer 14 times.
It was able to improve significantly when given access to the underlying filings. In “Oracle” mode, where it was pointed to the exact text for the answer, GPT-4-Turbo answered the question correctly 85% of the time, but still produced an incorrect answer 15% of the time.
But that’s an unrealistic test because it requires human input to find the exact pertinent place in the filing — the exact task that many hope that language models can address.
Llama2, an open-source AI model developed by Meta, had some of the worst “hallucinations,” producing wrong answers as much as 70% of the time, and correct answers only 19% of the time, when given access to an array of underlying documents.
Anthropic’s Claude2 performed well when given “long context,” where nearly the entire relevant SEC filing was included along with the question. It could answer 75% of the questions it was posed, gave the wrong answer for 21%, and failed to answer only 3%. GPT-4-Turbo also did well with long context, answering 79% of the questions correctly, and giving the wrong answer for 17% of them.
After running the tests, the cofounders were surprised about how poorly the models did — even when they were pointed to where the answers were.
“One surprising thing was just how often models refused to answer,” said Qian. “The refusal rate is really high, even when the answer is within the context and a human would be able to answer it.”
Even when the models performed well, though, they just weren’t good enough, Patronus AI found.
“There just is no margin for error that’s acceptable, because, especially in regulated industries, even if the model gets the answer wrong one out of 20 times, that’s still not high enough accuracy,” Qian said.
But the Patronus AI cofounders believe there’s huge potential for language models like GPT to help people in the finance industry — whether that’s analysts, or investors — if AI continues to improve.
“We definitely think that the results can be pretty promising,” said Kannappan. “Models will continue to get better over time. We’re very hopeful that in the long term, a lot of this can be automated. But today, you will definitely need to have at least a human in the loop to help support and guide whatever workflow you have.”
An OpenAI representative pointed to the company’s usage guidelines, which prohibit offering tailored financial advice using an OpenAI model without a qualified person reviewing the information, and require anyone using an OpenAI model in the financial industry to provide a disclaimer informing them that AI is being used and its limitations. OpenAI’s usage policies also say that OpenAI’s models are not fine-tuned to provide financial advice.
Meta did not immediately return a request for comment, and Anthropic didn’t immediately have a comment. More