CONTENTS. Forensic accounting: Identifying the Zone of Trouble..3. Methodology Accounting quality and investment returns 13

Transcription

1 STRATEGY December % 15% ne Zo 10% of Tro 5% Speculator Dodgy Auditor le ub Median share price performance Ambit Client 0% -5% D1 D2 D3 D4 D5 D6 D7 D8 Accounting score based deciles -10% D9 Cash Pilferage D10 Revenue Manipulation Forensic Accounting: Identifying the Zone of Trouble Analysts: Gaurav Mehta, CFA Tel: Karan Khanna Tel: Saurabh Mukherjea, CFA Tel:

2 CONTENTS Forensic accounting: Identifying the Zone of Trouble..3 Methodology Accounting quality and investment returns 13 Change in accounting quality and investment returns 18 Accounting quality in practice What does our model not capture? Sample bespoke - World Cargo December 22, 2014 Ambit Capital Pvt. Ltd. Page 2

3 THEMATIC December 22, 2014 Forensic accounting: Identifying the Zone of Trouble Forensic accounting is a key component of our research, as we have time and again shown that accounting quality is not just one of the many factors affecting investment returns but rather a critical hygiene factor, the lack of which can be detrimental to portfolio returns. Moreover, contrary to popular belief, accounting quality has stayed relevant even in the recent recovery, with stocks with weak accounting quality still underperforming. Through our well-established proprietary model, we seek to help investors avoid this accounting Zone of Trouble which is profoundly damaging for investment performance. Quantifying accounting quality Our model looks at the following key categories of accounting irregularities: balance sheet misstatement, profit & loss misstatement, cash pilferage and audit quality. We use 11 ratios across these categories to quantify accounting quality for stocks with a market-cap of more than Rs 1,000mn (excluding banks and financial services firms). The caveat is that whilst these aggressive accounting policies raise red flags, they may not necessarily imply accounting fraud. The Zone of Trouble Accounting quality is not just one of the many factors affecting investment returns but rather a critical hygiene factor, the lack of which can be detrimental to portfolio returns. An assessment of historical returns suggests that whilst the top 6 deciles on accounting do not seem to be materially different from each other on investment performance, the performance slumps beyond D6. Therefore, D7-D10 is the zone of trouble to be avoided at all costs. Accounting quality drives investment performance Median share price performance 25% 20% 15% 10% 5% 0% -5% -10% D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 Accounting score based deciles 'Zone of Trouble' Sector-neutral accounting buckets show a strong link between accounting quality and investment returns Accounting bucket Accounting score Share price performance Bucket A % Bucket B % Bucket C % Bucket D % Source: Bloomberg, Ambit Capital research Weakest ten sectors on accounting quality Sector Average accounting score Average share price performance Realty % Conglomerate % E&C % Telecom % Infrastructure % Miscellaneous % Capital Goods % Textiles % Utilities % Media % Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score is based on annual financials over FY09-14; stock price performance is sector CAGR from April 2008 to October THIS NOTE CANNOT BE USED BY THE MEDIA IN ANY SHAPE OR FORM WITHOUT PRIOR CONSENT FROM AMBIT CAPITAL. Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score is based on annual financials over FY09-14; stock price performance is from Apr 08 to Oct 14 (on a CAGR basis). Universe for this exhibit is BSE500. Furthermore, accounting quality has remained relevant even in the market recovery, with the last four deciles on accounting continuing to underperform significantly (in terms of investment returns). However, it is interesting to note that the biggest improvements in performance in the uptrend have come for D5 and D6 and not for firms with superior accounting quality (see pg 21). Are you in the Zone of Trouble? Our forensic accounting model allows us to conduct a first-level health check of portfolios and helps identify whether any holdings belongs to the Zone of Trouble. Further, on a bespoke basis for clients, we also supplement these screen-driven red flags with bottom-up investigative research on individual companies. Please contact your Ambit sales representative in case your portfolio has not been screened yet by our forensic accounting model. Analyst Details Gaurav Mehta, CFA gauravmehta@ambitcapital.com Karan Khanna karankhanna@ambitcapital.com Saurabh Mukherjea, CFA saurabhmukherjea@ambitcapital.com Ambit Capital and / or its affiliates do and seek to do business including investment banking with companies covered in its research reports. As a result, investors should be aware that Ambit Capital may have a conflict of interest that could affect the objectivity of this report. Investors should not consider this report as the only factor in making their investment decision.

4 Methodology We use 11 ratios to score our universe (excluding, banks and financial services firms) based on their accounting qualities. These ratios can broadly be categorised into four buckets. Exhibit 1: Key categories of accounting checks Category P&L misstatement checks Balance sheet misstatement checks Cash pilferage checks Audit quality checks Source: Ambit Capital research Ratios (1) CFO/EBITDA, (2) change in depreciation rate, and (3) volatility in Non-operating income (as a percentage of net revenues) (1) Cash yield, (2) change in reserves (excluding share premium) to net income excluding dividends, (3) provisions for doubtful debts as a proportion of debtors more than six months, and (4) contingent liability as a proportion of net worth (1) Non-operating expenses as a proportion of total revenues, (2) CWIP to gross block, and (3) cumulative CFO plus CFI to median revenues (1) CAGR in auditor s remuneration to CAGR in consol. revenues Here is a brief description of the accounting ratios with illustrative case studies: Note: These examples are just illustrative case studies. The objective is to demonstrate the working of our framework and not to imply any ill-intent on the part of the company in question. I - P&L misstatement checks 1 CFO/EBITDA: This ratio checks a company s ability to convert EBITDA (which can be relatively easily manipulated) into operating cash flow (which is more difficult to manipulate). A low ratio raises concerns about the company s revenue recognition policy (because this may imply aggressive revenue recognition through methods such as channel stuffing). We use a six-year median for this measure. Case study: Aurobindo Pharma Aurobindo s pre-tax CFO/EBITDA has consistently been significantly below the median for its peers. The company s CFO/EBITDA has deteriorated from 70% in FY10 to 46% in FY14 even whilst the CFO/EBITDA for its peer group has improved from 88% to 100% over the same period. This deterioration in its CFO/EBITDA is mainly on account of the deterioration in its working capital days (from 186 days in FY10 to 217 days in FY14). Given its lack of material exposure to branded generic markets, where terms of trade are more benign, its working capital days have remained at elevated levels throughout. We also note that in years when Aurobindo s EBITDA has posted healthy growth, CFO/EBITDA seems to have deteriorated, as working capital needs have expanded. For instance, whilst EBITDA improved by 17% in FY11, CFO/EBITDA deteriorated to 55% in FY11 vs 70% in FY10 owing to higher working capital requirements (204 days in FY11 vs 186 days in FY10). Note that inventories and debtors increased significantly during the year as well (32% and 29% respectively). Strategy We focus on four categories of accounting checks: P&L misstatement, balance sheet misstatement, cash pilferage and audit quality A low CFO/EBITDA may be indicative of aggressive revenue recognition policies December 22, 2014 Ambit Capital Pvt. Ltd. Page 4

5 Exhibit 2: Revenue recognition - Aurobindo Pharma vs its peers Company/Metric Pre-tax CFO as a % of EBITDA FY10 FY11 FY12 FY13 FY14 Average Aurobindo Pharma 70% 55% 65% 46% 46% 57% Cadila 88% 85% 64% 82% 98% 83% Glenmark N/A 60% 131% 80% 102% 93% IPCA 71% 74% 85% 77% 82% 78% Biocon 101% 150% 123% 104% 103% 116% Median (ex-aurobindo) 88% 80% 104% 81% 100% 88% Divergence (Aurobindo) -18% -25% -39% -35% -54% -32% Source: Company, Ambit Capital research 2 Change in depreciation rate: We calculate change in depreciation rates for each of the past six years (FY09-14). We then calculate the median of absolute changes and then sort the companies on this ratio such that the company with the smallest change in its depreciation rate receives the best score. The rationale is to penalise companies that have high volatility in their depreciation rate on a YoY basis. Case study: Amtek Auto Our model penalises high volatility in depreciation rate For the period under study, Amtek Auto uses the Straight Line Method to provide depreciation on its fixed assets in the manner and at the rates specified in Schedule XIV to the Companies Act This method and rate of providing depreciation are in line with that followed by most of the Indian companies. However, note that there was a significant YoY change in consolidated and standalone average depreciation rates in FY13. Whilst the acquisition of JMT Auto and Neumayer Tekfor Group (NT Group) in June 2013 could partially explain the volatility in depreciation rates for the year-ending September 2013 in the consolidated accounts, we do not understand why the YoY depreciation rates have declined even at the standalone level. Further, plant and machinery (subject to higher depreciation rates) constitute ~90% of Amtek Auto s gross block (vs ~78% for its peers). In spite of the higher share of plant and machinery in Amtek Auto s gross block, its depreciation rates have historically remained lower than its peers. Overall, Amtek Auto s lower depreciation rate vs its peers and the volatility in its depreciation rate suggest that further analysis would be warranted. Exhibit 3: Depreciation analysis - Amtek Auto vs its peers Company/metric Average depreciation rate YoY change in depreciation rate (bps) FY11 FY12 FY13 FY12 FY13 Amtek Auto (consolidated) 4.4% 5.0% 3.7% 66 (131) Amtek Auto (standalone) 4.7% 4.8% 3.7% 4 (103) Bharat Forge 6.7% 7.0% 6.6% 26 (41) Mahindra CIE 7.8% 7.6% 8.0% (12) 34 Nelcast 4.6% 4.6% 4.3% (3) (27) Median (ex-amtek) 6.7% 7.0% 6.6% (3) (27) Divergence (Amtek consolidated) -2.4% -2.0% -2.9% 69 (104) Divergence (Amtek standalone) -2.0% -2.2% -2.8% 7 (76) Source: Company, Ambit Capital research. Notes: (1) FY11 and FY12 figures for Amtek Auto are June year-end whilst FY13 is 15 months ended September 30, 2013 which has been annualised for a like-to-like comparison with peers. March year-end for all the other companies; (2) Standalone accounts for all the peer companies. 3 Volatility in non-operating income: We calculate change in non-operating income (as a percentage of net revenues) for each of the past six years (FY09-14). We then calculate the median of absolute changes and then sort firms on this ratio such that the company with the least volatility receives the best score. The High volatility in non-operating income is a cause of concern! December 22, 2014 Ambit Capital Pvt. Ltd. Page 5

6 rationale is to penalise firms where volatility in non-operating income is unusually high as this could imply intent to inflate profitability in years of low profits by resorting to such means as sale of assets, investments, and so on. Case study: Akzo Nobel Akzo Nobel s non-operating income (as a percentage of net revenues) has historically been higher vs its peers (see Exhibit 4 below). Akzo Nobel s NoI has averaged ~7% of net revenues over the last five years (vs ~1% for its paint sector peers). Further, this has also remained quite volatile over the period. For instance, NoI (as a % of net revenues) was 2.3% in FY14 v/s 6.3% in FY13 implying a ~394 bps YoY change. The high volatility in FY14 (~394bps) could be attributed to lower gains on redemption of long-term investments (in FY14 vs FY13) and interest on income tax refund in FY13. Exhibit 4: Akzo Nobel - Volatility in other income vs its peers Company/metric NoI as a % of net revenues Volatility in NoI (as a % of net revenues) FY10 FY11 FY12 FY13 FY14 FY11 FY12 FY13 FY14 Akzo Nobel 10.1% 9.1% 5.7% 6.3% 2.3% Asian Paints 2.0% 0.8% 1.1% 1.0% 1.1% Berger Paints 1.0% 1.3% 1.0% 0.9% 0.9% Kansai Nerolac 1.2% 2.1% 0.9% 0.6% 0.3% Median (ex-akzo Nobel) Divergence (Akzo Nobel) Source: Company, Ambit Capital research 1.2% 1.3% 1.0% 0.9% 0.9% % 7.8% 4.6% 5.4% 1.4% II - Balance sheet misstatement checks 4 Cash yield: This ratio is calculated as the yield earned on cash, investments and deposits. A low ratio could be a cause for concern, as it could mean that either the balance sheet has been misstated or that the cash is not being used in the best interests of the firm. We use a six-year median for this measure. Case study: Arshiya Historically, Arshiya appears to have earned a significantly lower yield on its cash and marketable investments as compared to its peers. Whilst its peers such as Gateway Distriparks have earned cash yields of 5-8%, Arshiya International s yields have been much lower. Exhibit 5: Arshiya - Cash yield vs its peers Company Name Investment income as a % of cash and marketable investments FY08 FY09 FY10 FY11 FY12 Arshiya International Gateway Distriparks Allcargo Logistics Source: Company, Ambit Capital research Whilst the company s cash yield did improve in FY13 (2.3%) and FY14 (3.2%), it is worth noting that in January 2013, several Arshiya employees made allegations of financial irregularities and claimed that the company had not paid salaries to its staff since September (Source: 09/news/ _1_lakh-shares-cent-promoters) A low cash yield may either imply balance sheet misstatement or that the cash is not being used in the firm s best interest December 22, 2014 Ambit Capital Pvt. Ltd. Page 6

7 5 Change in reserves (excluding share premium) to net income excluding dividends: This ratio is calculated by dividing the change in reserves (excluding share premium) on a YoY basis and dividing it by that year s PAT excluding dividends. We then take a six-year median of this ratio. A ratio of less than one indicates direct write-offs to equity without routing these through the Profit & Loss account and may indicate aggressive accounting policies. Case study: Tata Steel A ratio of less than one on change in reserves ex-share premium to net income exdividends may denote direct write-offs through the balance sheet Tata Steel acquired Corus (now Tata Steel Europe) in In FY08, the first year after consolidation, the defined benefit pension plan operated by Corus (British Steel Pension Scheme) had actuarial gains. These gains were recognised in the P&L (consistent with the Indian accounting standards) in the consolidated accounts of Tata Steel. From FY09, the company has changed its policy. Since then the pension liability of Tata Steel Europe Ltd has been computed and accounted for in accordance with IFRS and there have been actuarial losses from this defined benefit pension plan that do not hit the consolidated P&L. Instead, IFRS allows Tata Steel to take the actuarial losses through Reserves & Surplus. The justification provided by the company is that given the large share of Tata Steel Europe Ltd in the consolidated P&L, periodic changes in the assumptions underlying the computation of the pension liability would cause undue volatility in the stated consolidated profits. That said, we would like to highlight that the company has always made appropriate disclosures in its notes to accounts. This disclosure is also made by the company in its quarterly filings with the stock exchanges. We are highlighting Tata Steel as an example primarily to show how a few changes in accounting policies may cause a significant change to the reported bottom-line. In fact, the auditors of Tata Steel too have, since 2009, highlighted this in the audit report. For example, in the FY14 Annual Report, the auditors have made the following comments: Attention is invited to Note 42 to the financial statements regarding accounting policy for recognition of actuarial valuation change of Rs crores (net of taxes) [Gross: Rs crores] in the pension and other post retirement benefit plans of Tata Steel Europe Limited, a subsidiary for the reasons specified therein. Had the Company recognized actuarial valuation changes in the Consolidated Statement of Profit and Loss, the deferred tax expenses would have been lower by Rs crores and the Profit after taxes, minority interest and share of profits of associates would have been lower by Rs crores. Our opinion is not qualified in respect of this matter. If Tata Steel had followed the previous policy of recording actuarial valuation changes in the consolidated P&L, its restated earnings would have been different (as shown in the exhibit below). Especially interesting is the impact in FY09, the year in which Tata Steel Europe switched to the new method for estimating its pension liability. Exhibit 6: Tata Steel - Restated earnings if the previous policy had been followed (Fig in INR mn) FY08 FY09 FY10 FY11 FY12 FY13 FY14 Consolidated Profit after taxes [A] 123,500 49,509 (20,092) 89,827 53,898 (70,576) 35,949 Actuarial gain/ (loss) [B] 59070* (54,966) (35,412) (4,028) (23,723) (3,173) (6,282) Net profit (had the actuarial gains/(losses) been charged to the P&L) [C=A-B] 123,500 (5,457) (55,505) 85,799 30,175 (73,749) 29,667 Impact on net profit (as a % of stated profit) [(C-A)/A)] 0% -111% -176% -4% -44% -4% -17% Source: Company, Ambit Capital research, Note: * has already been reflected in the P&L. December 22, 2014 Ambit Capital Pvt. Ltd. Page 7

8 6 Provision for doubtful debts as a proportion of debtors more than six months: This ratio checks the conservativeness of a company s provisioning policy. A low ratio raises the spectre of earnings being boosted through aggressive provisioning practices. We use a six-year median for this measure. Case study: Jindal Steel and Power Historically, JSPL s standalone debtors outstanding for more than 6 months (as a percentage of gross debtors) have been 5-13%, slightly more than its steel peers. However, note that JSPL has provided for less than 1% of its debtors more than six months as of FY14. This is materially below the 60% figure for its steel peers. Even on a consolidated basis, JSPL s debtors outstanding for more than six months (as a percentage of gross debtors) have been 8-12%, slightly more than its power sector peers. In spite of a higher share of debtors outstanding for more than six months, on a consolidated basis, JSPL has provided for less than 1% of its debtors more than six months (vs ~60% by Tata Power in FY13 and FY14). Had the company s provisioning for debts more than six months been in line with its peers (i.e. 60%), consolidated PBT for FY14 would have been lower by ~5%. A low provisioning raises the spectre of earnings being boosted through aggressive provisioning practices Exhibit 7: Jindal Steel and Power - Provision for old debtors lower than its peers Company/metric Provision for doubtful debts as a % of debtors over six months Debtors over six months as a % of gross debtors FY11 FY12 FY13 FY14 FY11 FY12 FY13 FY14 JSPL (standalone) 4.3% 2.9% 1.4% 0.7% 3.9% 5.0% 6.7% 12.7% JSPL (consolidated) 4.3% 1.0% 0.8% 0.6% 2.5% 10.1% 8.4% 11.8% Steel players Tata Steel (standalone) 34.4% 32.9% 43.5% 63.0% 9.6% 3.8% 4.1% 3.2% SAIL (standalone) 59.3% 35.4% 27.4% 23.1% 5.6% 7.0% 10.0% 10.1% JSW Steel (standalone) 82.2% 39.8% 5.3% 61.1% 1.3% 1.5% 5.6% 21.1% Median (steel players) 59.3% 35.4% 27.4% 61.1% 5.6% 3.8% 5.6% 10.1% Divergence -55.0% -32.5% -26.1% -60.4% -1.7% 1.3% 1.1% 2.6% Power players NTPC (consolidated) 92.8% 88.7% 0.0% 0.0% 35.0% 12.6% 3.3% 6.8% Tata Power (consolidated) 75.9% 71.5% 56.5% 61.1% 9.0% 9.1% 10.4% 6.8% Median (Power sector players) 84.3% 80.1% 28.2% 30.6% 22.0% 10.9% 6.9% 6.8% Divergence (JSPL Consolidated vs Power peers) -80.0% -79.1% -27.4% -29.9% -19.5% -0.7% 1.5% 5.0% Source: Company, Ambit Capital research 7 Contingent liabilities as a proportion of net worth: This is a check on a company s off-balance-sheet liabilities. If this ratio is high, it raises concerns regarding the strength of the company s balance sheet in the event that the contingent liabilities materialise. Given that contingent liabilities also include genuine items such as letters of credit, bill discounting and capital commitments, we seek to eliminate as many of these items whilst computing the figure for contingent liabilities. We use a six-year median for this measure. Case study: Gillette India An example of a firm that gets penalized in our framework due to relatively high contingent liabilities as a proportion of net worth is Gillette India. FMCG companies in general have an asset-light model (implying relatively low networth) and hence this ratio might appear unusually high for some of these names. However, Gillette India has seen a meaningful rise in contingent liabilities over the last few years led by disputes around direct and indirect taxation matters, for which the company has chosen to appeal to the appropriate authorities. Contingent liabilities as a percentage of net worth have historically averaged ~45% for Gillette India (see Exhibit 8 below). Whilst contingent liabilities have historically remained high owing to disputes with respect to excise, service tax and customs matters, note that contingent liabilities in FY14 have increased A very high proportion of contingent liabilities to net worth indicates disproportionately high off-balance-sheet risk December 22, 2014 Ambit Capital Pvt. Ltd. Page 8

9 significantly on account of income tax disputes. This is with respect to matters related to income tax disputes on inventory write-off, allowability of losses carried forward from merged entities and others. Exhibit 8: Gillette India - Contingent liabilities analysis Company/metric Contingent liability as a % of net worth FY11 FY12 FY13 FY14 Gillette 26.4% 37.8% 43.8% 70.3% Dabur* 6.8% 7.5% 6.5% 8.9% Emami 17.1% 2.3% 8.8% 7.8% HUL 29.0% 21.7% 28.6% 28.0% Median (ex-gillette) 17.1% 7.5% 8.8% 8.9% Divergence (Gillette) 9.3% 30.3% 35.0% 61.4% Source: Company, Ambit Capital research. Note: *this does not include LCs, bills discounted. III - Cash pilferage checks 8 Non-operating expenses as a proportion of total revenues: This ratio checks a company s expenditure policy. A high ratio raises concerns regarding the authenticity of such expenses. We use a six-year median for this measure. Case study: Unity Infra Historically, Unity Infra s miscellaneous expenses (as a percentage of net sales) have been higher than its peers. Whilst this ratio has declined from ~1.5% in FY10 to ~0.9% in FY14, this is still well above the sector median of 0.14%. Higher miscellaneous expenses indicate less-than-ideal financial reporting practices and hence raise concerns regarding the authenticity of such expenses. Exhibit 9: Miscellaneous expenditure analysis - Unity Infra vs its peers Company/metric Miscellaneous expenses as a % of sales FY-11 FY-12 FY-13 FY-14 Average Unity Infra 1.49% 1.60% 1.01% 0.94% 1.26% Supreme Infra 0.21% 0.30% 0.42% 0.73% 0.41% MBL 0.83% 0.74% 0.89% 0.48% 0.73% JMC 0.14% 0.16% 0.06% 0.13% 0.12% Pratibha 0.21% 0.32% 0.11% 0.22% 0.22% KNR 0.07% 0.07% 0.11% 0.14% 0.10% J Kumar 0.19% 0.28% 0.51% 0.08% 0.27% Ahluwalia 0.06% 0.06% 0.04% 0.04% 0.05% Median (ex-unity Infra) 0.19% 0.28% 0.11% 0.14% 0.22% Divergence (Unity Infra) 1.30% 1.32% 0.90% 0.80% 1.04% Source: Company, Ambit Capital research 9 CWIP to gross block: The idea here is to penalise firms that show consistently high CWIP relative to the gross block, as this may either indicate unsubstantiated capital expenditure or a delay in commissioning (which may in turn be motivated by a delay in the recognition of the related depreciation expense). We calculate the proportion of capital work in progress to gross block for each of the last six years and then take the 25 th percentile observation (instead of a simple six-year median like in most other ratios). The reason for using the 25 th percentile over the last six years for this measure as opposed to the median (which would be the 50 th percentile observation) is to allow the benefit of doubt to firms that have invested wisely during the recent downturn. Hence we are penalising companies only if the ratio has been consistently high over most of the last six-year period. A high proportion of nonoperating expenses raises concerns regarding the genuineness of such expenses A high CWIP to gross block ratio may either indicate unsubstantiated capex or delay in commissioning December 22, 2014 Ambit Capital Pvt. Ltd. Page 9

10 Case study: Balkrishna Industries Balkrishna Industries is an example of a firm that gets penalised on this measure in our accounting model compared to its Auto Ancillary peers. Whilst the company s CWIP relative to its gross block was in line with its peers in FY11, the ratio has remained at elevated levels over the last three years. This seems to be on account of the capex incurred in connection with its green-field tyre project at Bhuj in Gujarat which has only been partly commissioned so far. As per the company, this plant will be fully commissioned by end-fy15. Hence, we believe CWIP (and thereby the CWIP/gross block ratio) should decline significantly in FY15. On bottom up coverage, we are BUYERs of the stock with a target price of Rs760 (27% upside). Exhibit 10: Balkrishna Industries - Capex analysis Company/metric CWIP/Gross Block FY11 FY12 FY13 FY14 Balkrishna Industries Apollo Tyres MRF N/A * JK Tyre Ceat Median (ex-balkrishna) Divergence (Balkrishna) (0.00) Source: Company, Ambit Capital research. Note: *N/A since MRF is a September ending company. 10 Cumulative CFO plus CFI to median revenues: We calculate the cumulative CFO (cash flow from operations) plus cumulative CFI (cash flow from investing activities) over the last six years and divide this by the last six-year median revenues for the company. The higher the ratio, the better our perception of the company s accounts. The idea is to penalise firms which over such large periods have been unable to either generate positive cash flows from operations or alternatively where cash flow from investments have consistently eaten away cash generated from operations. Case study: Amtek Auto Amtek Auto s standalone and consolidated gross block turnover has historically remained significantly below its peers. Plant and machinery have historically averaged ~90% of Amtek Auto s gross block (vs ~78% for its peers). Even with a higher share of Plant and machinery in overall gross block (which would imply a greater share of productive assets in its gross block versus its peers), Amtek Auto s gross block turnover hasn t picked up. Further, in spite of the lower gross block turnover, Amtek Auto s cumulative CFI over the last three years has consistently eaten away the cash generated from operations raising questions regarding the reasonability of its capex. Our model penalises firms that have not generated positive cash flows even on a six-year basis December 22, 2014 Ambit Capital Pvt. Ltd. Page 10

11 Exhibit 11: In spite of the lowest gross block turnover vs its peers Company/metric Gross Block Turnover FY11 FY12 FY13 Amtek Auto (consol) Amtek Auto (stan) Bharat Forge Mahindra CIE Nelcast Median (ex-amtek) Divergence (Amtek consolidated) Divergence (Amtek standalone) (0.6) (0.7) (0.8) (0.8) (0.8) (0.9) Source: Company, Ambit Capital research. Notes: (1) FY11 and FY12 figures for Amtek Auto are June year-end whilst FY13 is 15 months ended September 30, 2013 which has been annualised for a like-to-like comparison with peers. March year-end for all the other companies; (2) Standalone accounts for all the peer companies. IV - Audit quality checks Exhibit 12: Amtek Auto s free cash flows have been negative (Rs mn) Company/metric Cumulative CFO plus CFI Median revenues Cumulative CFO plus CFI to median revenues (FY11-13) (FY11-13) (FY11-13) Amtek Auto (consol) (85,847) 76,222 (1.1) Amtek Auto (stan) (19,601) 24,539 (0.8) Bharat Forge 3,841 31, Mahindra CIE 999 4, Nelcast 1,924 5, Median (ex-amtek) 0.2 Divergence (Amtek consolidated) Divergence (Amtek standalone) 11 CAGR in auditor s remuneration to CAGR in consolidated revenues: We calculate CAGR in standalone auditor s remuneration and consolidated revenues over FY A lower ratio of CAGR in auditor s remuneration relative to CAGR in consolidated revenues receives a high score. The rationale is to penalise companies whose growth in auditor s remuneration has exceeded the growth in the firm s revenues. Case study: Crompton Greaves Crompton s consolidated auditor s remuneration of Rs63.2mn in FY11 was far higher than the consolidated auditor s remuneration of Rs20mn paid by Infosys and standalone auditor s remuneration of Rs28.1mn paid by Larsen & Turbo in FY11. Further, whilst the company has stopped reporting consolidated audit fees since FY12 (as this was not required under the revised Schedule VI), Crompton s auditor s remuneration was significantly ahead of its peers whilst these numbers were still being reported. Exhibit 13: Crompton Greaves - Auditor s remuneration vis-à-vis peers (1.36) (1.03) Source: Company, Ambit Capital research. Notes: (1) FY11 and FY12 figures for Amtek Auto are June year-end whilst FY13 is 15 months ended September 30, 2013 which has been annualised for a like-to-like comparison with peers. March year-end for all the other companies; (2) Standalone accounts for all the peer companies. FY11 FY12 FY13 Crompton (consolidated) Net Sales (Rs mn) 100, , ,944 Auditor s remuneration (Rs mn) Not disclosed Auditor s remuneration as a % of Total revenues (%) Not available Siemens (standalone) Net Sales (Rs mn) 120, , ,526 Auditor s remuneration (Rs mn) Auditor s remuneration as a % of Total revenues (%) ABB India (consolidated) Net Sales (Rs mn) 62,871 74,518 76,105 Auditor s remuneration (Rs mn) Auditor s remuneration as a % of Total revenues (%) Alstom T&D India* (consolidated) Net Sales (Rs mn) 40,200 33,113 31,519 Auditor s remuneration (Rs mn) Auditor s remuneration as a % of Total revenues (%) Thermax (standalone) Net Sales (Rs mn) 48,524 53,041 46,909 Auditor s remuneration (Rs mn) Auditor s remuneration as a % of Total revenues (%) Source: Company, Ambit Capital research; Note: * FY12 numbers have been annualised. We penalise firms where growth in auditor s remuneration has been higher than the growth in the firm s revenues December 22, 2014 Ambit Capital Pvt. Ltd. Page 11

12 Cumulating scores: We cumulate scores across these 11 parameters to arrive at the final accounting score for each firm. Based on these parameters, we rank 370 BSE500 firms and 767 sub-bse500 firms on accounting quality in this year s forensic exercise. From the BSE500 universe, we have excluded 85 banks and financial services firms. Further, 8 firms were excluded because their FY14 annual reports had not been published at the time of running this exercise (see the table on the right). A further 37 firms have been excluded due to sketchy data availability/corporate restructuring/year-end change/limited listed history. Whilst we have extended this year s forensic accounting exercise to include all firms with mcap above Rs 1,000mn, the exhibits and discussion that you find in the subsequent sections are only for the BSE500 universe (excluding Financial Services firms). Data sources: We have used Ace Equity and Capitaline as data sources for the underlying financial data whilst stock price data has been sourced from Bloomberg. We had to use Ace Equity for some data items and Capitaline for some others in order to minimise data errors. Unfortunately, neither of these databases (nor any other database in India) is entirely reliable by itself. Please note, however, that several adjustments need to be made to each of the individual variables which we have not detailed here. For further details on these adjustments, kindly the authors of this note. List of firms whose FY14 data is not available (cut-off date is 31 October) Company name Trinity Tradeli. India Cements Sunrise Asian Alok Inds. Ballarpur Inds. Kappac Pharma Orchid Chemicals Luminaire Tech. Financial Year end March March March September June March September March Source: Ace Equity, Ambit Capital research Note: For the purpose of this exercise, we have included Essar Oil (March-ending), MRF, Siemens, Amtek Auto, Amtek India and BF Utilities (September-ending) based on their FY08-13 financials. December 22, 2014 Ambit Capital Pvt. Ltd. Page 12

13 Accounting quality and investment returns Absolute scores In this section, we seek to answer the following question: Does accounting quality, as quantified by our model, impact stockmarket performance? For this we assess the link between the blended forensic accounting scores for the BSE500 universe of firms, derived based on the methodology explained above using the last six years data, and the stock price performance for these stocks from April 2008 to October Universe level: We find that for the BSE500 universe as a whole, stock-specific accounting scores and stock price returns do not have a significant relationship (see the exhibit below). Such a lack of correlation is not surprising given the multitude of other factors (such as the underlying fundamental performance) which influence stock price returns. Exhibit 14: Scatter plot of accounting scores vs stock price performance for all BSE500 stocks does not bring out any significant relationship Share price performance 160% R² = 18% 140% 120% 100% 80% 60% 40% 20% 0% -20% % -60% Accounting score Stock-level noise leads to a weak relationship between accounting scores and stock returns at the universe level (370 firms) Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score is based on annual financials over FY09-14; stock price performance is from April 2008 to October 2014 on a CAGR basis. Universe for this exhibit is BSE Decile level: To control for noise around individual stocks, we arrange these stocks into deciles based on their accounting scores. We then find a strong relationship between the average accounting scores of these deciles and the average stock price performance of their constituent stocks, suggesting that accounting quality is a significant driver of stock returns. Exhibit 15: Decile-level analysis points to a strong link between accounting scores and stock price performance Median share price performance 30.0% 25.0% 20.0% 15.0% 10.0% -10.0% -15.0% R² = 86% 5.0% 0.0% D9 D7-5.0% D10 D8 D6 D5 D4 Accounting score based deciles D3 D2 D1 however, deciles constructed on accounting scores demonstrate the power of accounting quality in shaping stock returns Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accouting score is based on annual financials over FY09-14; stock price performance is from April 2008 to October 2014 on a CAGR basis. Universe for this exhibit is BSE500. December 22, 2014 Ambit Capital Pvt. Ltd. Page 13

14 In terms of individual decile performances, the first decile (D1) has delivered stock price returns of 24.2% CAGR since April In contrast, the last decile (D10) has delivered returns of -9.7% CAGR over this period, thus implying a close to 34% CAGR outperformance for D1 vs D10. The performance differential across deciles becomes more evident from the exhibit below. Exhibit 16: Decile-level analysis suggests accounting quality is important 25% Median share price performance 20% 15% 10% 5% 0% -5% 'Zone of Trouble' D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 Top accounting decile outperforms the bottom decile by 34% on a CAGR basis -10% Accounting score based deciles Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accouting score is based on annual financials over FY09-14; stock price performance is from April 2008 to October Shaded area denotes Zone of Trouble. Universe for this exhibit is BSE500. The most crucial takeaway from the above exhibit is that whilst the top 6 deciles on accounting do not seem to be materially different from each other on investment performance, the performance just slumps beyond D6. Therefore, thinking about accounting quality as just one of the many factors affecting investment returns isn t appropriate. It is in fact a critical hygiene factor, the lack of which can be seriously detrimental to portfolio returns; D7-D10 is the zone of trouble to be avoided at all costs. 3. Sector-agnostic buckets: One may argue that in the decile construction above, sector effects have not been nullified and some sectors may do better than others on our accounting model by virtue of the nature of their businesses. The decile performances thus might reflect serendipitous sector effects. To control for the sector effects, we now construct four sector-agnostic buckets such that bucket A comprises the first quartile of each sector on accounting scores, bucket B comprises the second quartile of each sector, bucket C comprises the third quartile of each sector and bucket D comprises the last quartile of each sector. Hence, every bucket has an equal number of stocks from each sector, implying that the buckets are sector-agnostic. Each bucket in this case will have similar sectoral compositions and hence a performance assessment of these buckets should enable one to assess the impact of accounting quality on stock price performance in a sector-agnostic manner. Exhibit 17 below displays these four buckets with their respective stock price performances. Clearly, the performance differential points to a strong link between accounting quality and stock price performances even after controlling for sector effects. December 22, 2014 Ambit Capital Pvt. Ltd. Page 14

15 Exhibit 17: Strong link between accounting quality and stock performance even after controlling for sector effects (the first entry is the accounting score over FY09-14, the second entry is the median CAGR stock returns in that bucket from April 2008 to October 2014) 25.0% 20.0% 15.0% 10.0% 5.0% 224, 19.2% 199, 14.3% 177, 11.8% 144, 2.0% Sector-agnostic buckets constructed with homogenous sectoral make and differentiated only on accounting quality show accounting quality drives investment performance even after controlling for sector effects 0.0% Bucket A Bucket B Bucket C Bucket D Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score is based on annual financials over FY09-14; stock price performance is from April 2008 to October Universe for this exhibit is BSE500. The above exhibit yet again highlights the vitality of avoiding the lowest rung firms on accounting for a healthy portfolio performance. 4. Sector level: Next, arranging the BSE500 firms into sectors and assessing the link between the average accounting scores of these sectors and the average stock price performance of their constituent stocks suggests that accounting quality makes a difference at the sector level as well (i.e. sectors with higher accounting quality, such as FMCG, Auto Anc and Consumer Durables, perform better than sectors with poor accounting quality such as Realty, Engineering & Construction and Infrastructure). Exhibit 18: Link between accounting quality and stock price performance at the sector level is relatively high Average share price performance 60% Consumer Durable 50% R² = 40% 40% Auto Agri 30% Chemicals inputs Auto Pharma Logistics IT Anc FMCG 20% Misc. Industrials Capital Textiles 10% Retail Conglomerate Media Cement goods E&C Oil & 0% Gas -10% Telecom Utilities Metals & Infra. -20% Mining Shipping Realty Average accounting score Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score is based on annual financials over FY09-14; stock price performance is from April 2008 to October Universe for this exhibit is BSE500. With an average score of 213, FMCG is amongst the best sectors in our accounting model. The sector has generated average stock price returns of 25% CAGR over the last six-year period since April On the other hand, Realty is the worst sector on accounting on our model with an average score of 152. The average stock price performance in the sector has been -12% CAGR over the last six-year period. Also, stocks within the same sector exhibit a significant link between accounting scores and stock price returns in many cases. Three sectors which show strong links are Utilities, Auto and Industrials. Link between accounting scores and price performance is relatively high at the sector level December 22, 2014 Ambit Capital Pvt. Ltd. Page 15

16 Exhibit 19: Within the sector, the link between accounting and price performance for Utilities 30% R² = 61% Share price performance 20% 10% 0% % -20% -30% Accounting score Even within a sector, stock returns show significant dependence on accounting scores Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score is based on annual financials over FY09-14; stock price performance is from April 2008 to October 2014 on a CAGR basis. Universe for this exhibit is stocks from the Utilities sector in the BSE500 index. Exhibit 20: Within the sector, the link between accounting and price performance for Auto Share price performance 90% R² = 21% 80% 70% 60% 50% 40% 30% 20% 10% 0% Accounting score Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score is based on annual financials over FY09-14; stock price performance is from April 2008 to October 2014 on a CAGR basis. Universe for this exhibit is stocks from the Auto sector in the BSE500 index. Exhibit 21: Within the sector, link between accounting and price performance for Industrials Share price performance 70% R² = 50% 60% 50% 40% 30% 20% 10% 0% -10% % Accounting score Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score is based on annual financials over FY09-14; stock price performance is from April 2008 to October 2014 on a CAGR basis. Universe for this exhibit is stocks from the Industrials sector in the BSE500 index. December 22, 2014 Ambit Capital Pvt. Ltd. Page 16

17 5. Size buckets: Finally to address the size dimension, we split our universe of stocks into four size buckets, as shown below. Bucket 1 comprises the largest 50 stocks on market cap, Bucket 2 of the next 100, Bucket 3 of the next 100 and Bucket 4 of the lowest 120 stocks on market cap (thus, taking the total to 370 firms). Accounting quality is better for larger-caps on average Exhibit 22: Larger capitalisation firms have better accounting scores on average Size Bucket Number of firms in the bucket Market cap range (Rs bn) Market cap range (US$ bn) Average accounting score Average share price performance % stocks in 'Zone of Trouble' Bucket 1 top 50 Rs331bn-5,108bn US$5.4bn-83bn % 24% Bucket 2 next 100 Rs68bn-330bn US$1.1bn-5.4bn % 31% Bucket 3 next 100 Rs28.4bn-66.8bn US$0.5bn-1.1bn % 37% Bucket 4 bottom 120 Rs4.4bn-28.3bn US$0.07bn-0.5bn % 56% Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score is based on annual financials over FY09-14; stock price performance is from April 2008 to October 2014 (on a CAGR basis). Universe for this exhibit is BSE500. As one would expect, we find that the average accounting score as well as the stock price performance varies directly with market cap, i.e. the larger market-cap buckets have better accounting scores as well as better stock price performance whilst the smallest market-cap bucket has the worst accounting score as well as the worst stock price performance. Further, the proportion of stocks in the Zone of Trouble (i.e. stocks that fall in D7 and below) too varies directly with market-cap. Whilst 24% of firms belonging to the largest market-cap bucket fall in the Zone of Trouble, ~56% of the firms belonging to the smallest market-cap bucket fall in this zone. Delving further into the stocks that fall in Bucket 4 of market-cap, we note that ~38% of these names also fall in Bucket D discussed earlier (i.e. the bottom quartile stocks from each sector). In contrast, the remaining ~62% stocks are evenly distributed amongst Buckets A to C (see Exhibit 23 below). Exhibit 23: Distribution of the smallest market-cap firms across the four accounting quality buckets Accounting quality bucket Number of firms As a % of total number of firms in Bucket '4' (i.e. the smallest 120 firms in the BSE500) Bucket A (best quality) 24 20% Bucket B 25 21% Bucket C 26 22% Bucket D (worst quality) 45 38% Total % Source: Company, Ambit Capital research. Universe for this exhibit is BSE500. December 22, 2014 Ambit Capital Pvt. Ltd. Page 17

18 Change in accounting quality and investment returns Change in scores The decile level and sector-agnostic buckets from the previous section suggest that accounting quality is a significant driver of stock prices and that this holds true even after controlling for sector effects. In this section, we seek to answer the question: Does a change in accounting score impact stockmarket performance? To calculate the change in accounting score, we break the six-year period (from FY09 to FY14) that we have used so far to calculate absolute accounting scores into two sub-periods FY09-11 and FY We then use the 11 parameters to quantify accounting scores for each of the two sub-periods separately using the same methodology as earlier (but for a three-year period now vs a six-year period earlier). Change in accounting score is calculated as the change in the FY12-14 sub-period s score over the FY09-11 sub-period s score. Finally, we assess the link between this change in accounting score for the BSE500 universe of firms and the stock price performance for these stocks from April 2011 to October Universe level: We find that for the BSE500 universe as a whole, the change in accounting scores and individual stock prices does not have a meaningful relationship (see the exhibit below). Such a lack of correlation is not surprising given the multitude of other factors (such as the underlying fundamental performance) which influence stock price returns. Exhibit 24: Scatter plot of change in accounting scores vs stock price performance for all BSE500 stocks does not bring out any significant relationship With accounting quality showing a strong link with stock price performance, change in accounting quality is another dimension meriting attention 250% R² = 0% Share price performance 200% 150% 100% 50% 0% (150) (100) (50) - -50% % Change in accounting score (FY12-14 over FY09-11) Again stock-level noise prevents any strong link between change in accounting scores and stock performance Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score change is for the FY12-14 subperiod over FY09-11; stock price performance is from April 2011 to October Universe for this exhibit is BSE Decile level: Similar to the methodology used in the preceding section to control for noise around individual stocks, we arrange these stocks into deciles based on their accounting scores. Arranging these stocks into deciles based on the change in accounting scores points to a moderate relationship between the change in accounting scores of these deciles and the average stock price performance of their constituent stocks. Decile-level analysis points to a moderate link between change in accounting scores and stock performance December 22, 2014 Ambit Capital Pvt. Ltd. Page 18

19 Exhibit 25: Decile-level analysis points to some link between the change in accounting scores and stock price performance but only a moderate one 30.0% Average share price performance 25.0% D6 D2 R² = 21% D8 20.0% D4 15.0% D1 D3 D7 10.0% D5 D10 D9 5.0% 0.0% (80) (60) (40) (20) Change in accounting score based deciles Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score change is for the FY12-14 subperiod over FY09-11; stock price performance is from April 2011 to October 2014 on a CAGR basis. Universe for this exhibit is BSE Market and sector level: When looking at changes in accounting scores over time, one is keen to know: (1) At the market level, are accounting ratios improving or worsening over time? (2) At the sector level, are accounting ratios improving or worsening over time? In the exhibit below, we highlight the proportion of ratios that are improving over time (i.e. in the FY12-14 period vs the FY09-11 period). It is heartening to see that on aggregate 80% of ratios have improved for India Inc. Exhibit 26: Improvement in accounting ratios at the overall market and sector level Sector Proportion of ratios improving (FY12-14 over FY09-11) Average stock price performance Aggregate 80% 15.5% Oil & Gas 90% 4.3% Pharma 90% 25.9% Auto Anc 90% 35.7% IT 80% 25.2% Logistics 80% 36.2% Realty 80% -2.1% Utilities 80% -3.5% Retail 80% 19.3% Infrastructure 70% 1.3% Cement 70% 28.3% Metals & Mining 70% -11.6% Conglomerate 70% 6.6% Industrials 60% 21.6% FMCG 60% 22.6% Auto 60% 33.0% Engineering & Construction 60% 1.4% Shipping 60% -0.8% Telecom 60% 3.7% Textiles 60% 10.0% Consumer Durable 60% 33.0% Agri Inputs 50% 29.9% Media 50% 17.8% Chemicals 50% 25.6% Capital Goods 50% 4.0% Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score change is for the FY12-14 subperiod overt FY09-11; stock price performance is from April 2011 to October 2014 on a CAGR basis. Universe for this exhibit is BSE500. Oil & Gas, Pharma and Auto Anc are amongst the most-improved sectors on accounting quality whilst improvements are more modest for Agri inputs, Media, Chemicals and Capital Goods December 22, 2014 Ambit Capital Pvt. Ltd. Page 19

20 At the sector level, the ratios of Oil & Gas, Pharma and Auto Anc have improved the most. Agri inputs, Media, Chemicals and Capital Goods bring up the rear of this table even with half of the parameters improving for these sectors. 4. Size buckets: Finally, we split our universe of stocks into four size buckets exactly in accordance with the method described in the preceding section. We find that the improvement in accounting scores is the most for the lower market-cap buckets. Exhibit 27: There does not appear to be link between capitalisation and change in accounting scores Number of firms in the bucket Size Bucket Market cap range (INR bn) Market cap range (USD bn) Proportion of ratios improving (FY12-14 over FY09-11) Average stock price performance top 50 Bucket 1 Rs 331bn-Rs 5,108bn US$ 5.4bn-US$83bn 40% 16.8% next 100 Bucket 2 Rs 68bn-Rs 330bn US$ 1.1bn-US$5.4bn 70% 17.9% next 100 Bucket 3 Rs 28.4bn-Rs 66.8n US$ 0.5bn-US$1.1bn 70% 24.3% bottom 120 Bucket 4 Rs 4.4bn-Rs 28.3bn US$ 0.07bn-US$0.5bn 90% 5.7% Source: Ace Equity, Capitaline, Bloomberg, Ambit Capital research; Note: Accounting score change is for the FY12-14 subperiod overt FY09-11; stock price performance is from April 2011 to October 2014 on a CAGR basis. Universe for this exhibit is BSE500. Overall, here are some of the key findings from our analysis of accounting quality change over time: At the universe level, the accounting quality of India Inc seems to be improving. At the sector level, Oil & Gas, Pharma and Auto Anc have improved the most. On the other hand, improvements have been more modest for Agri inputs, Media, Chemicals and Capital Goods Improvement in accounting ratios is more prominent for lower market-cap stocks. Improvement in accounting is the most for lower capitalisation stocks, helped by a lower base to start with December 22, 2014 Ambit Capital Pvt. Ltd. Page 20

21 Accounting quality in practice In this section of the note, we now turn to some practical observations on accounting quality based on our experience over the last few years. Accounting quality matters, even in a bull run That accounting quality drove alpha in the bear phase of is a point we have observed over the past few years. However, contrary to popular belief, accounting quality has remained relevant even in the recent uptrend, with the last four deciles on accounting continuing to underperform over the past 12 months. Exhibit 28: Lower deciles on accounting quality have underperformed even in the uptrend The Zone of Trouble on accounting quality has stayed an underperformer in the uptrend of the past 12 months 75% Median share price performance 60% 45% 30% 15% D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 Source: Bloomberg, Ambit Capital research. Note: Price performance has been measured over the period 21 November 2013 to 15 December However, it is interesting to note that in the uptrend the biggest improvements in performance have come for D5 and D6 and not for firms with champion accounting quality. Exhibit 29: Middle deciles have shown the best share price performance in the uptrend Stock price performance last 12 months* 12 months prior to that** Delta D1 48% 3% 45% D2 59% -4% 63% D3 44% 1% 44% D4 59% -7% 66% D5 64% -8% 72% D6 70% -10% 80% D7 44% -15% 59% D8 56% -19% 74% D9 31% -30% 61% D10 19% -33% 52% Average 49% -12% 62% Source: Bloomberg, Ambit Capital research. Note: * Defined as price performance over the period 21 November 2013 to 15 December 2014 ** Defined as price performance over the period 21 November 2012 to 21 November Accounting quality not priced in Satyam traded at a P/E discount to Infosys and Wipro even before the promoter owned up to aggressive accounting. Yet Satyam s share price crashed by over 90% within two days of the fraud being made public. Thus, the market does not already know and properly discount firms that have poor accounting quality. Accounting quality is not properly discounted by the markets December 22, 2014 Ambit Capital Pvt. Ltd. Page 21

22 We also find no correlation between P/E and accounting scores at the market level nor do we find anything significant at an intra-sector level. Accounting quality is not already priced in (and that is why it is worth investigating and assessing)! Exhibit 30: No correlation between accounting quality and P/E at the market level Exhibit 31: No correlation between accounting quality and P/E for FMCG stocks R² = 1% R² = 0% Trailing P/E Trailing P/E Accounting score Accounting score Source: Company, Ambit Capital research; Note: Trailing P/E has been restricted to 100. Universe for this exhibit is BSE500. Source: Company, Ambit Capital research; Note: Trailing P/E has been restricted to 100. Universe for this exhibit is stocks from the FMCG sector in BSE500 index. Not all Nifty firms have good-quality accounts Whilst the top size bucket, bucket 1 (Exhibit 22 on page 18), has the best accounting score, this overall average hides a great deal of variation even within the large caps. For example as many as 45% of Nifty firms have accounting scores below the universe (BSE500) average. Further, for the weakest nine of these firms, the accounting scores are actually so low that these firms fall in bottom four deciles of accounting quality for the BSE500. In other words, nine Nifty firms are in the Zone of Trouble. Avoiding the Zone of Trouble We can give interested clients an accounting heatmap of their portfolio within five working days of receiving it if the constituent stocks are in our accounting model. This will enable clients to identify if any of their holdings are in the Zone of Trouble. A sample screenshot of what such a diagnostic looks like is presented below. Exhibit 32: Indicative portfolio heatmap Companies Ambit sector CFO Cont Liab- EBITDA % of NW Change in depr rate Vol. in NoI (as a % of sales) CAGR in auditor s remn to CAGR in consol revs Scores Non-oper exps-% of total revs Cash yield PFD-% of debtors more than six months Cum. FCF/ Change in median reserves/(pat revs ex dividend) ABC Industrials XYZ Utilities PPP Utilities DEF Metals GHI Metals RRR IT TTT Oil & Gas PQR Oil & Gas Note: ORANGE denotes sub-par accounting quality relative to the sector average; Red denotes that the stock falls in the Zone of Trouble For a more detailed analysis, we also do extensive bottom-up company-specific bespoke research for clients. A sample bespoke has been attached towards the end of this note (page 27 onwards) as an example of our past work on this front. Not all firms from the Nifty have clean accounting Overall Score December 22, 2014 Ambit Capital Pvt. Ltd. Page 22

23 What does our model not capture? Whilst our accounting model uses 11 objective, quantifiable parameters to assess the overall quality of a firm s accounts, it has certain limitations too. For example, there are certain subjective assessments on governance standards that one can make only after going through the annual reports and which cannot be quantified. Our bottomup analysis therefore plays a vital role in helping our clients identify these parameters. Some such examples are discussed below. Subjective checks- Related party transactions Investors should lookout for suspicious related party transactions undertaken by listed entities. At its simplest, the extent of related party transactions and the trend in these transactions are important; a sudden increase can be bad news. However, it is often trickier than this, as shown in instances illustrated here. Parties would be considered related if at any time during the financial year, one party is able to either control the other party or can exercise significant influence over the other. Thus, related parties would include subsidiaries, associates, joint ventures, key management personnel and their relatives, etc. Ideally, transactions between related parties should be at arm s length. An arm s length transaction would mean that both the parties seek to execute the transaction in their best interests. However, in several cases, related party transactions are conducted in a manner that is not in the best interests of one party. Overpaying for an asset purchased from a related party, sale of goods or other assets to related parties at a significant discount to their fair market values, loans given to related parties at exceptionally concessional rates or loans taken from related parties at exorbitant interest rates are just a few examples of how these transactions might not be in the best interests of the minority shareholders. Likewise, unwarranted transactions with related parties should raise a red flag. This point can be better understood by analysing certain related party transactions that Crompton Greaves has undertaken over the last 5-6 years. (Note: Crompton Greaves falls in the second quartile in its sector on our accounting model). Case study: Crompton Greaves During FY08, Crompton Greaves purchased co-ownership rights in an aircraft from a related party, M/s Asia Aviation Ltd, for `562.5mn. Mr. Gautam Thapar, MD & CEO of Crompton Greaves, was also a director of M/s Asia Aviation Ltd, a company in the business of providing air charter services. Whilst it is arguable that the aircraft purchase was unwarranted, the fact that it had been executed with a related party in the business of providing aircrafts on a lease basis also raises concerns regarding the appropriateness of such a transaction, given that Crompton Greaves could have simply hired the aircraft. This transaction was followed by the purchase of another aircraft during FY11 for `2700mn. However, no disclosures were made by the company in its annual report for FY11 as to whether or not this was a related party transaction. When these issues were raised by investors with the management in 2QFY12, the management transferred the entire block of aircrafts at book value to its unlisted related parties, M/s Asia Aviation Ltd (`411.7mn) and M/s Avantha Holdings Ltd (`2,405mn). This last point can be detected from the FY12 Annual Report. Subjective checks - Issues raised by the auditor Amongst the first things to look for whilst analysing an auditor s report is to keep an eye out for any qualifications made by the statutory auditor. Whilst there are several reasons why an auditor may qualify his reports, one common reason is when the accounts have not been drawn up according to the generally accepted accounting principles. Suspicious related party transactions would merit further attention Investors should keep an eye out for any qualifications made by the statutory auditor December 22, 2014 Ambit Capital Pvt. Ltd. Page 23

24 This can be better understood by looking at the issues raised by Lanco Infratech s auditors, PwC, in the auditor s report in FY07 and FY08. (Note: Lanco Infratech falls in the bottom quartile in its sector on our accounting model). Case study: Lanco Infratech In FY07, the auditors of Lanco Infra had raised the following issues with respect to the consolidated financial statements: Profits were higher by `169.29mn (or 9% of consolidated profits for FY07) due to non-elimination of intra-group transactions and unrealised profits pending clarification from ICAI. The consolidated financial statements were presented considering M/s Lanco Kondapalli Power Private Limited (LKPPL) as a subsidiary with effect from 1 April 2006 when in fact, LKPPL became a subsidiary of the company with effect from 15 November As a result, profits were higher by `242.94mn (or 13% of consolidated profits in FY07). Not only did the company not meet the requirements of AS-21 on the consolidated financial statements (given that according to AS-21 issued by the ICAI, consolidation should have been carried out from 15 November 2006, the date on which holding subsidiary relationship came into existence), the above treatment resulted in the profits for FY07 being higher by `412.23mn (or 22% of consolidated profits for that year). The excess profit of `412.23mn, which was recognised in the P&L account of Lanco Infra in FY07, had to be reversed in FY08. According to the generally accepted accounting principles, such a reversal should have been made against current-year profits (i.e. FY08 profits in this case) as a prior-period adjustment. Lanco Infra instead chose to adjust these excess profits against the balance of profit brought forward from the previous year. Consequently, the correction to FY08 profits was not made as required. The auditors too had raised their issues on such a treatment in their report on the consolidated financial statements for FY08. Issues raised by Lanco s auditors in FY07 Annual Report (on Page 63) Attention is drawn to the following: As detailed in note 4(xvi) of Schedule 19, pending clarification from the ICAI on non-elimination of intra group transactions and unrealized profits arising out of construction of projects under Build Operate Own and Transfer basis, the Company has not eliminated revenues and unrealized profits in the consolidated financial statements. As a result the consolidated revenue and net profit after minority interest are higher by ` millions and ` millions respectively. M/s Lanco Kondapalli Power Private Limited (LKPPL) has become a subsidiary of the Company with effect from November 15, However the consolidated financial statements have been presented considering LKPPL as a subsidiary with effect from April 01, As a result the consolidated revenues and net profit after minority interest are higher by ` and ` millions respectively. Issues raised by Lanco s auditors in FY08 Annual Report (on Page 71) Attention is drawn to Note 4.viii on Schedule 19 to the consolidated financial statements regarding the adjustment of excess profits recognised in the previous year aggregating to ` million against the balance of profit brought forward from the previous year, which in our opinion and according to the generally accepted accounting principles in India should have been adjusted against the current year s profit as a prior period adjustment. Consequently the net profit after tax and minority interest for the current year has been overstated by the above amount. December 22, 2014 Ambit Capital Pvt. Ltd. Page 24

25 Subjective checks - Cash repatriation to foreign parent Ideally, cash held by the subsidiary belongs as much to the minority shareholders as it does to the controlling shareholder. However, in recent times, we have seen how the parent can abuse its position as the dominant shareholder to pull out cash from the subsidiary without the minority shareholders having any say in this decision. In order to protect the interests of minority shareholders, SEBI came out with a Circular in early February (which was later amended in May), wherein it directed that listed entities would need the support of majority of the minority shareholders in case of M&A transactions. Ambuja Cement is a recent example of a firm that has seen repatriation of cash to the foreign parent. (Note: Ambuja Cement falls in the top quartile in its sector on our accounting model). Case study: Ambuja Cement What was proposed? In July 2013, the Board of Directors of Ambuja Cement unanimously approved the proposal to amalgamate Holcim (India) Private Limited (HIPL) with the company. HIPL is a wholly owned subsidiary of Holderind Investments Ltd, Mauritius. As a result of the scheme, Ambuja acquired a 50.01% stake in ACC and paid `35bn in cash and issued incremental 433.7mn shares worth `83bn for the same to Holcim (see Exhibits 33 and 34). Exhibit 33: From a complicated structure earlier Cash repatriation to foreign parent is another aspect that investors should watch out for Exhibit 34: Holcim transfers all its holdings to Ambuja Source: Company, Ambit Capital research Source: Company, Ambit Capital research Issues with the proposal As highlighted by our Cement analyst and Head of Research, Nitin Bhasin, in his 25 July 2013 note (click here for the detailed note), the rearrangement would not result in any value creation for Ambuja s shareholders because: The controlling shareholders stood to benefit at the expense of the minority shareholders. The rearrangement offered a relatively raw deal for minority shareholders, as it would result in an indirect holding in ACC. It is difficult to decipher how the transaction in which the cash is taken from Ambuja and replaced with investment in ACC made the capital structure of the standalone Ambuja entity any better. Holcim took out `35bn cash giving in return only the promise of some future synergies. The synergy potential of `9bn do not seem likely to accrue in the near future (see Exhibit 35). Exhibit 35: Expected synergies and benefits highlighted by Ambuja Cement s management to be extracted over the next two years Source Supply Chain Optimisation Shared services / Fixed Costs Total Synergy/ cost benefits Through `3.6bn-4.2bn Clinker swaps, cement swaps Procurement; fixed cost reduction `4.2bn-4.8bn through shared services in back-end processes; financial optimisation `7.8bn-9bn Source: Company, Ambit Capital research Our view This is a likely source of immediate benefit and with the change in the management structure, these benefits could accrue soon; we expect benefits of this in CY14 itself Our primary data sources suggest that both the companies are already working on most of these together especially procurement and hence we do not expect such large benefits immediately We model total benefits of `4.5bn in CY14, equally for ACC and Ambuja and then increasing marginally every year December 22, 2014 Ambit Capital Pvt. Ltd. Page 25