People in the real world often possess vague knowledge of future payoffs, for which quantification is not feasible or desirable. We argue that language, with differing ability to convey vague information, plays an important but less-known role in representing subjective expectations. Empirically, we find that in their reports, analysts include useful information in linguistic expressions but not numerical forecasts. Specifically, the textual tone of analyst reports has predictive power for forecast errors and subsequent revisions in numerical forecasts, and this relation becomes stronger when analyst's language is vaguer, when uncertainty is higher, and when analysts are busier. Overall, our theory and evidence suggest that some useful information is vaguely known and only communicated through language.

Apple reported better-than-expected earnings in the third quarter of 2024, with buzz about its new AI features offsetting a continuing decline in its key China market. Earnings exceeded analyst predictions despite a year-over-year decline in iPhone sales, with revenue rising 4.9% to 85.78bn in the three months ending 29 June, beating the average analyst estimate of 84.53bn. The company maintained its cash dividend at 25 cents for each share. The strong report represented a bright spot in the tech space after difficult earnings reports from other tech giants like Amazon, Snap and Intel. The market saw a sell-off on Thursday amid disappointing results, including from Intel – which announced plans to cut more than 15,000 jobs as it tries to cut billions of dollars in costs to turn its business around.

There is considerable evidence that machine learning algorithms have better predictive abilities than humans in various financial settings. But, the literature has not tested whether these algorithmic predictions are more rational than human predictions. We study the predictions of corporate earnings from several algorithms, notably linear regressions and a popular algorithm called Gradient Boosted Regression Trees (GBRT). On average, GBRT outperformed both linear regressions and human stock analysts, but it still overreacted to news and did not satisfy rational expectation as normally defined. By reducing the learning rate, the magnitude of overreaction can be minimized, but it comes with the cost of poorer out-of-sample prediction accuracy. Human stock analysts who have been trained in machine learning methods overreact less than traditionally trained analysts. Additionally, stock analyst predictions reflect information not otherwise available to machine algorithms.

Tokyo stocks gained Wednesday, led by issues of companies that released their earnings reports the day before. The 225-issue Nikkei average climbed 62.08 points, or 0.21%, to end at 29,053.97, after shedding 134.34 points Tuesday. The Topix index of all first section issues advanced 5.51 points, or 0.29%, to finish at 1,909.06, following a 14.60-point drop the previous day. The Tokyo market began on a weak note as players took to position-squaring selling ahead of the Golden Week holiday period, starting Thursday. Stocks headed north later in the morning thanks to buying of firms which released their earnings results and forecasts Tuesday such as industrial robot-maker Fanuc and electronics maker Fuji Electric.

Equity research analysts at financial institutions play a pivotal role in capital markets; they provide an efficient conduit between investors and companies' management and facilitate the efficient flow of information from companies, promoting functional and liquid markets. However, previous research in the academic finance and behavioral economics communities has found that analysts' estimates of future company earnings and other financial quantities can be affected by a number of behavioral, incentive-based and discriminatory biases and systematic errors, which can detrimentally affect both investors and public companies. We propose a Bayesian latent variable model for analysts' systematic errors and biases which we use to generate a robust bias-adjusted consensus estimate of company earnings. Experiments using historical earnings estimates data show that our model is more accurate than the consensus average of estimates and other related approaches.

Artificial intelligence training provider Appen issued a strong earnings upgrade on Monday morning on the back of growing business from its existing customers and signs that the woes have subsided at its recent acquisition, Figure Eight. The company, which makes money by crowdsourcing labour for artificial intelligence/machine learning services for tech giants such as Google, said full-year earnings before interest, tax, depreciation and amortisation (EBITDA) is expected to be in the range of $96 million to $99 million. If current forex rates hold, this could add a further $1.5 million to EBITDA this year. "Appen's improved FY2019 earnings forecast is driven by increases in monthly relevance revenues and margins, largely from existing projects with existing customers," the company said.

This quarterly earnings report is not subject to the quarterly review procedures in accordance with the Financial Instruments and Exchange Act. At the time of disclosure of this quarterly earnings report, the review procedures for quarterly financial statements in accordance with the Financial Instruments and Exchange Act are incomplete. The earnings forecasts are based on information currently available to the Company at the time of the release of these materials. Actual business results may differ from the forecasts due to various factors. For information regarding the assumptions on which earnings forecasts are based and points to note when using the earnings forecasts, please refer to "(3) Information regarding consolidated earnings forecasts and other forward-looking statements" under "1. Supplementary material on quarterly earnings will be available on the Company's website, on Friday, January 27, 2017.