.png)
Investing in publicly listed companies carries inherent risks, one of which is the potential for accounting manipulation. Transparently.AI's Risk Engine (TRE) is designed to help investors assess this risk more effectively. This report focuses on how the TRE works and how investors can use it to make more informed investment decisions.
Our fundamental premise is this: All things being equal, fund managers who manage to eliminate companies with worse-than-average accounting quality from their portfolio will markedly improve their chances of outperforming the market.
Why accounting manipulation impacts stock returns
First we need to understand why that is so, and our starting point here is a clear acknowledgment that any attempt at account manipulation, such as premature revenue recognition or the improper capitalization of expense, will always negatively impact future earnings. Anything that impacts future earnings - by even a few percentage points - will necessarily impact future stock returns.
This is an important point of understanding and is worth fleshing out. For a fuller treatment of this, we have a specific article, but for this article, let’s consider two examples:
A company seeks to boost current earnings by delaying essential expenses: This creates an illusion of profitability in the current reporting period that cannot be maintained because the company will need to incur these expenses in future financial periods. Moreover, the more a company delays essential opex, such as maintenance, the greater the risk of a disruption to future business activity or a decline in product quality. Both will affect future profits.
A company improperly capitalizes expenses: When a company capitalizes an expense, it records the expense as an asset and must depreciate or amortize the cost of the asset over future years. This is appropriate for long-lived assets that will give service over time, but not appropriate for operating expenses and SG&A expenses.
These should be recognized in the income statement in the period when they occur. The more aggressively a company capitalizes operating expenses, thus delaying expense recognition, the more amortization costs will be in future years. Profit margins must necessarily fall in future years. Companies that aggressively delay expense recognition create a ticking expense time bomb that eventually comes back to haunt them.
Manipulating accounting figures creates a misleading picture of a company's financial health. This can boost stock prices or secure favorable financing in the short term. However, these actions are unsustainable and will lead to a deterioration in earnings or impair solvency in the long-term.
This underpins the correlation between accounting quality and stock returns, which is more than adequately demonstrated in two extensive white papers prepared by the Transparently team. This relationship highlights the importance of incorporating accounting risk assessment into investment strategies.
Higher accounting quality means superior stock performance
The simplest demonstration of the powerful link between accounting quality (as measured by the Transparently Risk Engine) and subsequent stock returns is shown in Figure 1, which is taken from one of the white papers.
This chart is constructed using a dataset composed of all stocks listed globally anytime from January 2000 until October 2024, excluding banks and insurers, stocks with less than 3 years of financial data and any stocks with insufficient financial data to populate the TRE. This results in a total of 61,306 unique companies and approximately 7.5 million months of observations; quite possibly the most extensive stock investigation ever undertaken.
In the construction of Figure 1, the TRE was employed to generate risk scores (accounting quality measures) for each company at the end of each month. These scores were calculated as point-in-time estimates, meaning that only data available at month end was employed and the entire system was re-estimated each month on those respective datasets. This simulates true historical values for risk scores from the TRE.
Each month, companies were sorted into deciles based on their risk scores, with decile 1 containing companies with the lowest risk, and decile 10 containing companies with the highest risk. A selection of companies identified as “known manipulators” were eliminated from the data set to remove any model bias since these companies are important for model training.
The median future 12-month return was calculated for each decile over the entire period. These median decile returns are presented in Figure 1.
Figure I: Median 12-month future returns for portfolios formed on TRE risk deciles
Source: Transparently.AI
In Figure 1, we can see 12-month future median returns (absolute and relative) across the full universe and time period investigated for each decile of the Transparently risk score. We can clearly see the strong negative relationship between score decile and future 12-month returns. Also note that this is a non-linear relationship, with high risk score deciles associated with progressively stronger implications for future returns. The difference in the future returns between the decile for lowest risk and that for highest risk is 26.7%.
In the white paper, this effect was examined for return periods as short as a month and as long as 36 months. The difference in future returns increased from 3.3% over 1 month periods to 52.6% over 36 month periods.
The bottom line is that investors should absolutely consider incorporating measures of accounting quality into their investment processes. Professional investors desiring to investigate the system can request the white papers from our sales consultants.
We are not advocating the TRE as a stand-alone stock selection tool. Many factors influence stock returns and there is always the possibility that poorly run companies can turn around. Nevertheless, the Transparently.AI can greatly augment the investment process.
What is the Transparently Risk Engine?
The TRE leverages advanced artificial intelligence (AI) and forensic accounting techniques to analyze publicly available data and identify potential red flags in a company's financial reporting. The system assigns a risk score to each company, ranging from 0 to 100%, with 0 indicating high quality and 100 indicating low quality.
In addition to the unique scoring system, the TRE embodies four key features:
- Dynamic risk assessment: The TRE constantly updates its assessments based on new information, such as global company announcements, ensuring that investors have access to the most up-to-date risk profiles.
- Unique risk factor combinations: Unlike traditional static models, the TRE identifies unique combinations of risk factors for each company and year, providing a more nuanced and accurate assessment.
- Cluster and factor analysis: The TRE aggregates hundreds of component models into 14 different accounting risk clusters to identify specific areas of concern within a company's financial reporting.
- Transparently ratings: In addition to the risk score, the TRE provides a Transparently Rating from A+ to F, which combines the risk score with relative comparisons within the world, sector, and year.
The system can provide a state-of-the-art assessment of any company’s accounting quality within seconds. It takes a team of forensic analysts weeks, if not months, to do the same thing.
How investors can use the TRE
The TRE can be integrated into various stages of the investment process:
- Screening and due diligence: Investors can use the TRE to screen potential investments and identify companies with high accounting manipulation risk. This allows for a more focused due diligence process, concentrating on companies with potentially higher risk.
- Portfolio risk management: By incorporating TRE data into their risk management frameworks, investors can gain a better understanding of the overall risk profile of their portfolios and make more informed decisions about asset allocation and diversification.
- Engagement and monitoring: The TRE can be used to identify companies that may require further engagement or monitoring due to potential accounting irregularities. This information can be used to initiate dialogues with company management or adjust investment strategies accordingly.
The TRE can provide investors with valuable insights, such as identifying companies with unusual trends in key financial ratios, which may indicate earnings manipulation; highlighting companies with aggressive revenue recognition practices or unusual accounting policies; and detecting companies with weak corporate governance structures or a history of auditor disagreements.
The TRE can answer a wide range of questions that are critical for investors, including: What is the overall accounting manipulation risk score for a specific company? How has the risk score for a company changed over time? What are the top risk factors and clusters contributing to a company's risk score? How does a company's risk score compare to its peers in the same sector and region? What is the estimated potential financial impact of accounting manipulation on a company?
The type of question that the system can answer is really only limited by your imagination. Top-down investors, for example, can use the TRE to understand differences in accounting quality across different sectors or even countries, and to gain insights into trends in the same. Deteriorating accounting quality, such as occurred in several countries in the 1990s, can warn of rising systemic market risk.
Conclusion
The TRE is a powerful tool that can help investors navigate the complexities of financial reporting and make more informed investment decisions. By providing a data-driven assessment of accounting manipulation risk, the TRE empowers investors to identify potential red flags, manage portfolio risk, and ultimately, generate higher portfolio returns.
%20-%20Wirecard.png)
