Each individual stock is evaluated on its own merits and its relative attractiveness to its peers. This analysis offers a deep understanding of the market and dynamically reacts to market changes as they happen. Deshe’s analysis methodology is as follows: 

  1. Each parameter is modeled against historical (15-25 years) & recent performance, tailored to its sector/industry benchmark 
  2. The momentum & rate of change is calculated for each parameter
  3. Unique “Deshe-made” distributions & standard deviations are calculated and attributed per parameter, relative to their peer group
  4. Based on the company’s sector and industry, Deshe’s algorithms identify and choose the most relevant & effective parameters, derivatives, and ratios for predicting stock performance for each specific stock analyzed 
  5. Analyze The resulting data is analyzed using a stochastic methodology, based on  historical performance
  6. Deshe’s proprietary Machine Learning techniques are applied
  7. The probability assessment & final grading of expected performance per stock is calculated to optimize stock selection for the relevant timeframe
  8. Each company is provided with a numeric score, reflecting their likelihood of overperforming their peer group (for example 8/10)
  9. Stocks are designated with a relative numeric rank, translated into actionable insight

"Strong Buy" (5 Stars) and "Buy" (4 Stars) ratings represent the highest statistical likelihood of a company to overperform in its sector & industry.

"Hold" (3 & 2 Stars) rating indicates the company is not statistically likely to overperform.

"Underperform" (1 Star) rated companies are the most statistically unlikely to overperform their peers.

Deshe Analytics rating is an AI-driven projection/opinion and not a statement of fact. Any action in any security is subject to personal research and risk assessment.

Yes. We are fully transparent on that. In each report, you can see the weight allocation for each parameter and section (i.e balance sheet weight and revenue momentum weight). The weight is presented within the text paragraph in each report section. 



We cover ~37,000 public companies from all global markets. Our coverage threshold is $10MM market-cap, on the latest financial statement publish date.

Based on the daily data inflows related to market volatility, the company- & benchmark-specific updates, and price fluctuations, we measure the accuracy of our assessments and update our algorithms and rankings accordingly. In most cases, the fundamental analysis will not change dramatically until the next company's financial statements.

For example, a company received a “Buy” ranking right after their filing date and since then the stock price has appreciated 10% over the benchmark. Today, when I look at the attractiveness of the company, Deshe’s algorithm will have updated the relative attractiveness of this company’s upside and downside potential. It is likely that the company's ranking will be updated to “Hold” or even “Underperform” on the date the query is made. This makes sense because, at today’s updated price, the stock may have a lower upside potential / higher risk of price decrease relative to its peer group.

  1. The company didn’t publish new financial statements in the last 6 quarters
  2. On the latest financial statement publish date, the company market cap was less than $10MM
  3. The company went public during the last 2 quarters and didn’t publish 2 financial reports yet

We look at the companies from the same industry and exchange. Then, we divide the companies into two different MarketCap groups - companies larger than $300MM and companies smaller than $300M.

Yes! Within each sector, we provide an identical number of ratings for Buys and Underperforms, according to the following chart:

Rank Percentile

Rating

100-95

Strong Buy

95-80

Buy

80-20

Hold

20-0

Underperform 

Yes. For each sector & industry, we run different analysis algorithms. The algorithms identify and choose the most relevant & effective parameters, derivatives, and ratios for predicting stock performance for each specific stock analyzed, in order to maximize the success rates of the analysis for each specific market segment.

Yes. We differentiate the analysis algorithms for different sectors and industries, adapting the algorithm based on the varieties of geographical location and market-cap. 

In most cases, we are providing the analysis within 12-48 hours from the financial statement release. There are 2 major exceptions:

  1. During the 1st and 2nd week of the reports season, we may postpone some of the analysis updates for specific companies because our analysis is comparative,  i.e. we need a minimum threshold for companies' reports within the relevant sector in order to have enough comparative data. The existing threshold is 2%.
  2. In rare cases, where we identify extreme and abnormal deviations from the previous report analysis & peers' behavior, we postpone the update in order for our product team to review the data and confirm everything is OK.

Please note that the majority of Deshe’s analysis is based on the full financial statements (AKA: 10K, 10Q, etc.). Once that financial statement is published, we are usually providing the analysis within 12-48 hours from the financial statement release. There are 2 major exceptions:

  1. During the 1st and 2nd week of the reports season, we may postpone some of the analysis updates for specific companies - because our analysis is comparative - we need a minimum threshold for companies' reports within the relevant sector in order to have enough comparative data. The existing threshold is 2%.
  2. In rare cases, where we identify extreme and abnormal deviations from the previous report analysis & peers' behavior - we postpone the update in order for our product team to review the data and confirm everything is OK. 

Also - there are exceptions for companies that Deshe is not analyzing (3 main reasons):

  1. The company didn’t publish new financial statements in the last 6 quarters
  2. On the latest financial statement publish date, the company market cap was less than $10MM
  3. The company went public during the last 2 quarters and didn’t publish 2 financial reports yet

Our algorithms analyze companies’ YoY trailing 12 months fundamental data. In order for the analysis to follow our quality specifications, we must measure the change and rate of change for each financial parameter, and compare those to the analyzed company’s peers. Therefore, all the companies we analyze must have a minimum of 2 financial statements published.



No. All the rankings and ratings are following the Deshe Analytics methodology with no external interventions and are 100% data-driven.

Deshe has developed a unique NLG (Natural Language Generation) process. Deshe’s NLG is a multi-step process designed to produce content with natural-sounding speech/text. 

In general, there are six primary components of Deshe’s NLG automated writing process:

  1. Content Analysis: This stage identifies the major themes of the source document and the relationships between them. Here, the data is analyzed and filtered to determine what exactly should be included in the content produced at the end of the process. 
  2. Understanding the Data: Machine learning is often used in this phase. The data is interpreted, patterns are identified, and they are placed into context. 
  3. Structuring the Document: A report plan is created and a narrative structure is chosen based on the type of conclusion needed.
  4. Aggregation of Sentences: Relevant sentences or fragments of sentences are combined to accurately summarize the topic.
  5. Structuring based on Grammar: The syntax and structure of the target sentence are identified. Thereafter, grammatical rules are applied to rewrite the sentence appropriately, with the end result being a human-sounding, grammatically correct text.
  6. Presentation of the Text: The final output is generated based on a template or format selected by the user or programmer/designer. 

In the report’s top section, we show the last trading day's close price (In black). Right below the price we show the PPS on the financial statement filing date:

Alphabet