The purpose of this monthly article is to provide investors with additional tools to evaluate financial irregularities according to three different detection models from academic research. It is my hope that this contribution toward improving due diligence in the market place using publicly available data and published algorithms is of some value to the reader. Circumstances surrounding firms are always subject to change, open to extenuating circumstances, and models by their very nature always contain a degree of error.
I learned of these financial detection algorithms and others, during my doctoral research on financial anomalies as well as through texts on financial statement irregularities in my certification programs. Again, it is important to understand that any firm identified by these academic models may not be in actual distress or suffer from any adverse irregularities whatsoever. These models are certainly not foolproof and were designed by academic researchers to improve the chance of detection of irregularities leading to bankruptcy, earnings manipulation, or flag the presence of financial distress.
To my knowledge no similar study of positive and adverse irregularities using these three models has been conducted before. I hope that benefits are also constructively considered regarding why such anomalies may exist in these models at this moment in time. The resulting data which varies from month to month may prompt firms and investors to consider further due diligence of publicly available financial characteristics to mitigate any risk or error present in the marketplace.
The process for screening stocks begins by applying the positive and adverse parameters for each of the three academic models: In the adverse selection, stocks are screened for the combination of values less than 1.81 on the Altman Z-score, greater than 50% on the Ohlson O-score probability, and greater than -2.22 on the Beneish M-score model. In the positive selection, stocks are screened for the combination of values greater than 2.99 on the Altman Z-score, less than 10% on the Ohlson O-score probability, and less than -4.44 on the Beneish M-score model.
The Altman (1968) Z-score was developed by Edward Altman as a bankruptcy prediction model for "practical decision making." He broke from traditional ratio analysis and introduced a unique approach leveraging multiple discriminant analysis (MDA) to determine which financial ratios provided the most effective signals of financial distress when used simultaneously. His work with MDA was also inspiration for my ongoing research towards identifying the strongest characteristics of price momentum behaviors.
The five financial ratios that Altman relied upon are: working capital/total assets, retained earnings/total assets, earnings before interest and tax/total assets, market value of equity/book value of total liabilities, and sales/total assets. Within his sample of 66 firms, the model had a 95% accuracy in classifying corporate failure within a one-year event horizon.
The parameters used for Altman Z-score span the values from the safe zone > 2.99 to the risk zone < 1.81.
The Ohlson (1980) O-score probability was developed by James Ohlson and used a multi-factor financial model to predict business failure that relied on nine financial ratios. Olson used a sample of 105 bankrupt firms and 2,058 non-failing firms to classify event failures with financial ratios. The results did not perform as well as MDA, but well enough to contribute additional high probability of distress detection.
The nine financial ratios are: firm size, total liabilities/total assets, working capital/total assets, current liabilities/current assets, a dummy variable whether total assets were greater or less than total liabilities, net income/total assets, funds from operation/total liabilities, a final dummy variable whether net income was negative for the last two years, and change of net income.
The parameters used for the Ohlson O-score probability set that firms with probabilities above 50% would likely default within two years, while values less than 10% represent strong underlying fundamentals.
The Beneish (1999) M-score was developed by Messod Beneish and used a weighted blend of eight different indexes to measure year-over-year changes in order to detect earnings manipulations. This scoring method was also accurate in anticipating the high risk of earnings manipulation prior to Enron's demise. The eight ratios used are: Days' sales in receivables index (a measure of current year over prior year sales and receivables), gross margin index (a ratio of prior year's gross margin to the current year), asset quality index (a ratio of current year's non-current assets other than P&E to total assets over the prior year's value), sales growth index (a ratio of current year's sales to prior year's sales), depreciation index (a ratio of prior year rate of depreciation over current year depreciation rate), SG&A expense index (a ratio of current year over prior year's sales, general, and administrative expenses), leverage index (a ratio of current year over prior year's total debt to total asset ratio), and lastly the total accruals to total assets ratio.
The parameters used for the Beneish M-score set a benchmark of at-risk firms having a value greater than -2.22, while values less than -4.44 show a low risk of earnings manipulations or irregularities.
The July Stock Screen Results
The July stock screen was run for both the positive and adverse parameters of each of the three models and results were obtained using a proprietary database at UncleStock.com. In the positive selection screen, the database returned 41 stocks from among global stock exchanges not including OTC stocks. Only seven stocks (17%) on the list were both US-based and priced above $3 per share. These stocks are shown in the table below.
In the adverse selection screen the database returned 131 stocks from among the primary stock exchanges not including OTC stocks. Stocks were sorted in descending order based on last traded stock price from highest to lowest. Only 11 stocks (8.4%) of the list had share prices above $3. In both screens, the stocks that lacked the requisite data to return meaningful results from all three applied models were removed from the list. So both resulting lists represent the stocks above $3 per share that had scores for Altman Z, Ohlson O, and Beneish M corresponding to the positive and adverse benchmarks for each parameter.
The seven stocks within the favorable parameters of all three detection algorithms for July are John Bean Technologies (JBT), Spark Therapeutics (ONCE), Northwest Pipe Company (NWPX), AC Immune (ACIU), Pure Cycle (PCYO), Friedman Industries (FRD ), and R1 RCM, Inc. (OTCPK:RCM). The average YTD performance of this portfolio is 9.54% and may be indicative of undervalued fundamental characteristics.
In the adverse parameters of all three detection algorithms for July are the following seven stocks: Omeros Corporation (OMER), China Online Education Group (COE), Immunomedics (IMMU), XOMA Corporation (XOMA), root9B Technologies (OTC:RTNB), ImmunoGen (IMGN), ViewRay (VRAY). The average YTD performance of this portfolio is 78.55% and may be indicative that high performance characteristics tend to skew fundamental values measured by the three models in this study.
Benchmark share prices for future reference: OMER $21.96, COE $17.32, IMMU $8.66, XOMA $7.27, RTNB $7.08 , IMGN $5.66, VRAY $5.17
I leave the interpretation of these results to the reader. The YTD stock market performance differences between the two selection groups are significant and may provide you with a new way to identify value and growth selections in the stock market. Time will tell how significant these very different outcomes may be, but for now this study provides a small baseline from which to consider the fundamental selection criteria using a combination of Altman Z, Ohlson O, and Beneish M scores.
The goal of this article is to raise awareness about financial anomalies in the marketplace using publicly available data and published academic algorithms. My ongoing research on accelerating breakout and breakdown momentum seeks to identify irregularities in whatever form they may arise. Circumstances surrounding firms are always subject to change, open to many extenuating circumstances, and models by their very nature always contain a degree of error. Publicly available databases are not without error and due diligence should always be applied before jumping to any conclusions from any of the models offered in this article.
It is my sincere hope this research is of some value to you as a practical guide in the marketplace applying more tools from well known academic methods. I hope you will join me in observing what results emerge from this monthly study.
Altman, E. I. (1968). The Prediction of Corporate Bankruptcy: A Discriminant Analysis. The Journal of Finance, 23(1), 193–194. doi:10.1111/j.1540-6261.1968.tb03007.x
Beneish, M. D. (1999). The Detection of Earnings Manipulation. Financial Analysts Journal, 55(5), 24–36. doi:10.2469/faj.v55.n5.2296
Ohlson, J. A. (1980). Financial Ratios and the Probabilistic Prediction of Bankruptcy. Journal of Accounting Research, 18(1), 109. doi:10.2307/2490395
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Editor's Note: This article covers one or more stocks trading at less than $1 per share and/or with less than a $100 million market cap. Please be aware of the risks associated with these stocks.