In a Nutshell

Deshe’s Equity Analysis Insights reporting combines cutting-edge machine learning techniques tailored to equity markets, artificial Intelligence algorithms, and a pinch of style. We instantly create fundamental reports in any language for 44,000 global equities. Our reports are compressed to the essentials and therefore make for an easy read. The process of creating reports for each of the 44,000 global equities is built on 5 pillars:

Deshe Report Mockyp - Transparent (1)
  • 1 Big (Raw) Data Harvesting
  • 2 Establish Benchmarks
  • 3 Stock Analysis Methodology
  • 4 Daily, Constant Supervision
  • 5 NLG - Easy Reading
data

Big (Raw) Data Harvesting

 

Deshe harvests historical and live market data for +44,000 publicly traded companies from the most reliable 3rd party data sources (ie. via API from S&P). This raw data includes +280 fundamental parameters and factors (balance sheet, cash flow, income statement, and further relevant ratios and multiples) pulled from company filings (quarterly financials) and 10K reports (annual financials).

In addition, non-fundamental alternative data is sourced from 3rd party data providers. This data includes insider trading (​​Form-3), changes in institutional ownership (13F), news reports, earnings reports & calls, analyst coverage, etc.

Group 512

Prioritization of comparative benchmarks, broken down by parameter/factor

 

Peer groups are based on the company's GICS (Global Industry Classification Standard) defined sector and industry. Each company is analyzed and evaluated against its relevant peer group.

Each company is then analyzed for its past performance, the performance is then ranked by individual factor, and each company’s future performance is dynamically forecasted in real time.

Individual factor and parameter benchmarks are calculated and applied, adjusted to their relevant sector and industry, and updated constantly as new information is generated from live market performance.

Group 500

44.000 stocks analyzed and evaluated against their peers

 

Deshe provides quarterly analysis and coverage for more than 44,000 individual global stocks. Each individual stock is evaluated on its own merits and its relative attractiveness to its peers. This analysis deeply understands the market and reacts dynamically to market changes.

Deshe’s methodology for analysis is summarized as follows:
Each parameter is modeled against historical (15-25 years) & recent performance, tailored to its sector/industry benchmark.

Momentum & rate of change is calculated for each parameter.
Deshe’s proprietary distribution & standard deviations are calculated and attributed to each parameter relative to their peer-group.

Based on the company’s sector and industry, our algorithms identify and choose for each analyzed stock the most relevant & effective parameters, derivatives, and ratios to predict stock performance. The resulting data is analyzed using a stochastic methodology based on historical performance.

Deshe’s proprietary Machine Learning techniques are applied
Probability assessment & final grading of expected performance is calculated per stock to optimize stock selection for the relevant timeframe.

A numeric score is assigned to each company, reflecting their likelihood of overperforming their peer group (for example 8/10)
A relative numeric rank is assigned to each stock, which translates into actionable insight.

Group 508

Constant supervision and dynamic updating of data and benchmarks

 

Our AI never sleeps and constantly receives market inputs from various financial, news and data sources to provide an accurate and up-to-date assessment for each equity.

Dynamic market data and market dynamics are included in the ongoing assessment and include among others market volatility, company news, benchmark specific updates and price fluctuations.

All these inputs are fed back into updating our algorithm to ensure the most accurate model that reflects current market situations.

Group 381

Translating quantitative information into clear, human language insights

 

Not everyone prefers graphs and charts to straightforward human language. For investors that want the clarity of a written explanation, Deshe developed an NLG ("Natural Language Generation) engine to translate our quantitative, numeric analysis into meaningful written insights. We are committed to helping investors simply and easily glean the most important & actionable information about the stocks they are interested in. Moreover, we do this in multiple languages.

 
Big (Raw) Data Harvesting
data

Big (Raw) Data Harvesting

 

Deshe harvests historical and live market data for +44,000 publicly traded companies from the most reliable 3rd party data sources (ie. via API from S&P). This raw data includes +280 fundamental parameters and factors (balance sheet, cash flow, income statement, and further relevant ratios and multiples) pulled from company filings (quarterly financials) and 10K reports (annual financials).

In addition, non-fundamental alternative data is sourced from 3rd party data providers. This data includes insider trading (​​Form-3), changes in institutional ownership (13F), news reports, earnings reports & calls, analyst coverage, etc.

Establish Benchmarks
Group 512

Prioritization of comparative benchmarks, broken down by parameter/factor

 

Peer groups are based on the company's GICS (Global Industry Classification Standard) defined sector and industry. Each company is analyzed and evaluated against its relevant peer group.

Each company is then analyzed for its past performance, the performance is then ranked by individual factor, and each company’s future performance is dynamically forecasted in real time.

Individual factor and parameter benchmarks are calculated and applied, adjusted to their relevant sector and industry, and updated constantly as new information is generated from live market performance.

Stock Analysis Methodology
Group 500

44.000 stocks analyzed and evaluated against their peers

 

Deshe provides quarterly analysis and coverage for more than 44,000 individual global stocks. Each individual stock is evaluated on its own merits and its relative attractiveness to its peers. This analysis deeply understands the market and reacts dynamically to market changes.

Deshe’s methodology for analysis is summarized as follows:
Each parameter is modeled against historical (15-25 years) & recent performance, tailored to its sector/industry benchmark.

Momentum & rate of change is calculated for each parameter.
Deshe’s proprietary distribution & standard deviations are calculated and attributed to each parameter relative to their peer-group.

Based on the company’s sector and industry, our algorithms identify and choose for each analyzed stock the most relevant & effective parameters, derivatives, and ratios to predict stock performance. The resulting data is analyzed using a stochastic methodology based on historical performance.

Deshe’s proprietary Machine Learning techniques are applied
Probability assessment & final grading of expected performance is calculated per stock to optimize stock selection for the relevant timeframe.

A numeric score is assigned to each company, reflecting their likelihood of overperforming their peer group (for example 8/10)
A relative numeric rank is assigned to each stock, which translates into actionable insight.

Daily, Constant Supervision
Group 508

Constant supervision and dynamic updating of data and benchmarks

 

Our AI never sleeps and constantly receives market inputs from various financial, news and data sources to provide an accurate and up-to-date assessment for each equity.

Dynamic market data and market dynamics are included in the ongoing assessment and include among others market volatility, company news, benchmark specific updates and price fluctuations.

All these inputs are fed back into updating our algorithm to ensure the most accurate model that reflects current market situations.

NLG - Easy Reading
Group 381

Translating quantitative information into clear, human language insights

 

Not everyone prefers graphs and charts to straightforward human language. For investors that want the clarity of a written explanation, Deshe developed an NLG ("Natural Language Generation) engine to translate our quantitative, numeric analysis into meaningful written insights. We are committed to helping investors simply and easily glean the most important & actionable information about the stocks they are interested in. Moreover, we do this in multiple languages.

 
70.4%

Of Strong Buy’s overperformed their benchmarks in 2021

+7.53%

Average over-performance of Strong Buys compared to benchmarks (2020-2021)

91%

Of our analysis algorithms overperformed their benchmarks (2020-2021)

-5.48%

Average performance of “Sell”s compared to benchmarks (2020-2021)

 

*Past performance is no guarantee of future results.

More visibility = More opportunities