We live in the Internet era where information is in surplus. The same applies to the context of retail investors in the stock market, who stand inundated in a flood of news, data and analyses. Practically, no one has the time and wherewithal of someone like Michael Burry, to pore over megabytes of data and uncover a massive collapse like the global financial crisis, before it unravels. Similarly, in a busy life, an ordinary investor cannot practically perform a forensic balance-sheet deep dive for every company in their portfolio. The sheer volume of data makes thorough due diligence nearly impossible.
This is where the Beneish M-Score steps in as an effective, quick-screening tool to detect frauds. Developed by Professor Messod Beneish, this mathematical model crunches key financial ratios to provide a preliminary assessment of whether a company is potentially manipulating its financial records.
Having written on the M-score a couple of times this year (in bl.portfolio editions dated March 2 and April 20, 2025), you might wonder why we are writing on this again. Well, on December 3, Kotak Institutional Equities (Kotak IE) put out a note on Kaynes Technology India (Kaynes), after analysing the company’s FY25 annual report. It had multiple allegations of accounting irregularities, giving one the idea that the books are likely to have been manipulated. This sent the stock price tumbling 18 per cent over Thursday and Friday. The company has since filed a document clarifying the allegations in a stock exchange filing. More on this below.
This piqued our curiosity to run the M-score on Kaynes. Kaynes failed the test, with a score of -1.1, which means the company is likely to have been engaged in earnings manipulation (see chart). However it is important to note here that M-score can give false alarms, and hence is only a tool to filter and research further on companies. The M-score formula, the formulae for the ratios or indices that make the formula, and the final findings have all been given in charts.
In the case of Kaynes are these results conclusive? Let’s find out.
The score
DSRI stands for days sales in receivables index, and it can also be looked at in tandem with SGI. Beneish believed companies might go lax on credit policy (customers payback in how many days) to inflate sales figures, ultimately delaying cash flow. Kaynes’ receivable days saw only a slight uptick from 72 days in FY24 to 77 days in FY25, rendering DSRI output as 0.99 (77/72 x coefficient 0.92).
GMI stands for gross margin index. The logic behind this ratio is similar to that of DSRI. Beneish thought that in a bid to boost revenue, companies tend to go easy on pricing, impacting their gross margin. A lower margin in the current year versus the base year might mean risk of manipulation. In Kaynes’ case, FY25 gross margin at 30 per cent is in fact, higher than FY24’s 26 per cent, rendering GMI output as 0.46 (26/30 x coefficient 0.528).
AQI stands for asset quality index. Beneish’s formula considers non-current assets except plant, property, equipment (PPE) to be not of ‘quality’ because he believed that such assets are generally prone to cost deferrals. Cost deferral is when a company decides to capitalise a revenue expense. In Kaynes’ case, the proportion of non-current assets except PPE (we have also added non-current investments to PPE, as they are largely quoted mutual fund units) in FY25 was 29 per cent of total assets versus 9 per cent in FY24. This 29 per cent over 9 per cent multiplied by the coefficient 0.404 rendered the index’s value as 1.31. This is primarily on account of a 5.2x jump in other non-current assets. Again, this jump is largely due to a trade receivables item of ₹326 crore in FY25, which was nil in FY24 (note this, as we refer to this later).
SGI stands for sales growth index, which basically measures revenue growth in the current year over the base year. Beneish believed that companies tend to manipulate revenue growth when under pressure from investors to deliver on targets, via tactics including round tripping. Kaynes’ revenue grew 51 per cent in FY25 or 1.51 times FY24 sales. SGI works out to 1.35 on multiplying 1.51 by the coefficient 0.892.
DEPI stands for depreciation index. Per the model, it is likely a case of manipulation if the current year’s depreciation rate is lower than that of the base year. This is because, companies tend to tweak estimates such as useful life, residual value of assets to their advantage to keep depreciation expense down and eventually inflate earnings. Rate of depreciation and amortisation has decreased in FY25 from 7.6 per cent to 5.1 per cent, suggestive of potential manipulation. But the result of this index at 0.17, in the overall M-score stands diluted because of its low coefficient of 0.115.
SGAI stands for sales, general, administration expenses index. Beneish theorised that in genuine cases of revenue growth, SGA expenses also rise commensurately. The opposite would indicate risk of manipulation. In Kaynes’ case, SGA expenses as a proportion of revenue have risen with revenue in FY25, at about 10 per cent versus 8 per cent in FY24, rendering the outcome as -0.21 (10/8 x coefficient -0.172).
LVGI stands for leverage index. Beneish believed that companies that struggle with cash flow, rely more on external funding and that increases leverage. Higher leverage in the current year versus base year suggests manipulation is unlikely, as higher debt indicates the company’s recognition of a cash crunch. Beneish sees this as a positive from manipulation standpoint, as more debt invites scrutiny from lenders, credit rating agencies and the like. Kaynes’ leverage ratio (including lease liabilities), has risen in FY25 at 20 per cent of total assets versus FY24’s 10 per cent, yielding the value -0.64 for the ratio.
TATA stands for total accruals to total assets. Accruals refer to the difference between a company’s accounting profit and cash profit, and this is divided by total assets for the purpose of comparison with other firms. The ability to convert accounting profits to cash profits reflects the true quality of earnings. A high delta between the two could signal a risk of manipulation, where the company is keen on propping up accounting profit than realised profit. In FY25, while Kaynes made a profit of ₹293 crore, its operating cash flow was a negative ₹82 crore, driven by increase in working capital, which in turn is mainly due to the other non-current assets of ₹326 crore, mentioned earlier. Even with the highest coefficient among the eight ratios, the result of this ratio works out to 0.38, failing to move the score as strongly as SGI, for instance, towards manipulation territory.
Adding all ratios together leaves one with a M-score of -1.1. But as seen above, the ratios are a mixed bag — some indicate risk of manipulation, while some don’t. Here, we wish to remind readers that the Beneish M-score is not bullet-proof. It offers a good starting point for filtering companies but doesn’t offer a definite conclusion, when used as a standalone measure.
What Kotak IE said
In its note, Kotak IE had pointed multiple irregularities, some of which have been clarified by the company in an exchange filing on Friday. However, what forms the eye of the storm is a company named Iskraemeco India Private Ltd (IIPL), which Kaynes acquired in September 2024 for ₹43 crore. IIPL is primarily into the business of designing smart meters. But it also bids at tenders floated by power discoms for installation of smart meters. It does not manufacture the meters, but buys all of its requirement from Kaynes.
In FY25, Kaynes’ revenue grew 51 per cent over FY24. Kotak IE believes this acquisition to have played a material role in its growth. In its separate financial statements, IIPL revenue in FY24 was ₹65 crore. In just a year, this jumped 9.5 times to ₹618 crore in FY25. In H2 FY25 – the period during which Kaynes consolidated IIPL’s books, consolidated revenue grew 44 per cent over FY24 or ₹500 crore. It appears that ₹175 crore (based on Kotak IE’s assumed net margin of 28 per cent for IIPL) of this ₹500 crore or about 35 per cent is attributable to IIPL . Similarly, consolidated net profit for H2 FY25 grew ₹56 crore over FY24 or 44 per cent. Of this, ₹49 crore (given in annual report) is attributable to IIPL, meaning net profit of Kaynes’ existing businesses (ex-IIPL) in H2 FY25 grew barely 6 per cent, casting a doubt on their performance.
While this definitely rises eye brows, based on digging into past management conference calls and other publicly available information, we were able to arrive at some more information.
When companies bid and win smart meter contracts from discoms, it becomes public information on this website — https://www.nsgm.gov.in/en/sm-stats-all. According to this web site, by FY24 close, IIPL would have had an order-book to deploy 37 lakh meters in West Bengal and 35 lakh meters in Gujarat. Deployment has begun in the months of July 2023 and January 2024 (in FY24) respectively. Based on a rough estimate of ₹3,000 revenue per meter, which the management cited in a conference call in October 2024, IIPL would have deployed about 2.1-2.2 lakh meters in FY24 (both States combined). However, per the web site, IIPL has deployed 11.9 lakh meters in Gujarat and 5.3 lakh meters in West Bengal cumulatively. From this, one can understand that deployment would have ramped up faster in H2 FY25, relative to FY24. While this is a plausible explanation for IIPL’s high revenue growth in FY25, the fact that Kaynes’ other businesses grew merely at 6 per cent remains unclear.
Moving on, we ask readers to recall the ₹326 crore of receivables from AQI. This is housed within other non-current assets. Kotak IE points out that this played its part in the cash conversion cycle getting extended. In its entirety, this appears to belong to IIPL only and the same was consolidated with the parent. Going by the Q4 FY25 earnings call transcript, it comes across that these receivables relate to a certain software revenue stream, which IIPL could collect from its clients over really extended periods. Since they aren’t technically due within the immediate working capital cycle, IIPL has accounted them under non-current assets.
Now Kaynes is in the process of ‘bill discounting’ these receivables with financiers, which has already been done to some extent. However, the arrangement is on a ‘with recourse’ basis, meaning if IIPL’s customers do not honour the claim when they become payable, the financier can claim the same from Kaynes. The management has, in fact, resolved to bring the credit period down to between 90 and 120 days, prospectively for new contracts. However, in so far as the receivables that it has already discounted, per accounting standards, it must recognise a contingent liability (also pointed out by Kotak IE), because these arrangements are on a ‘with recourse’ basis.
While discounting can augment cash flows in the short term, whether IIPL’s clients pay up when the time comes, needs a close watch.
The above are plausible explanations we have gathered from publicly available data. However, there are other grey areas and what we need now is convincing explanations.
IIPL’s contribution to group profit of ₹49 crore in H2 FY25 implies an unrealistic payback period of just six months (remember Kaynes bought IIPL for ₹43 crore). The small acquisition price sounds too good to be true, when you consider IIPL had already won valuable orders in West Bengal and Gujarat. Why did the seller of the company dump it at a fire sale price? Two, the then deputy CFO and CEO resigned from the company in January and October 2025, respectively, both citing personal reasons. This is unusual. Three, Promoter and Managing Director Ramesh Kunhikannan sold 11.25 lakh shares in June in a bulk deal at about ₹5,500 per share, representing about 2 per cent of outstanding shares then.
In the light of recent allegations, these make one wonder whether these are coincidental or consequential. The management’s integrity appears to have taken a beating. It is fair play by the company clarifying the irregularities on Friday. However, there is still some ambiguity as highlighted above and now the ball is squarely in the management’s court to come up with more detailed responses on its own. If done, it would greatly help allay investors’ concerns and also keep speculation at bay.
For a company trading at PE of 77 times even after recent correction, the M-score alone must suffice for investors to stay away and avoid bottom fishing.
Published on December 7, 2025
