Angel Tax for Start-ups - DCF creating a hullaballoo?

February 14,2019
Rate this story:
Smarak Swain, IRS

With rising volume of transactions in unlisted securities – in the form of PE investing, mergers & acquisitions, corporate restructuring etc – company valuations are proving to be consequential in corporate business decisions. Business valuations also form key documentation in fulfilling regulatory criteria appurtenant to such transactions. 

Although there are multiple methods using different degree of mathematical rigours, the Discounted Cash Flow(DCF) method is most prevalent and widely accepted in the corporate world. In this method, the future cash flows arising from the business’ present assets are calculated and future cash flows are discounted to present day to find the Enterprise Value (EV). Not only is DCF a favourite among valuers, it is also prescribed as most appropriate method for finding the fair market value of shares of unlisted companies under the Foreign Exchange Management Act (FEMA) and the Income Tax Act.

FEMA mandates that any transfer of shares from a resident to a non-resident should not be at a price less than the fair value of such shares, and any transfer of shares from a non-resident to a resident should not be at a price more than the fair value. DCF method was a mandatory method for finding the ‘fair value’ under the FEMA till July 2014. Vide Circular No. 4/2014 dtd 15.07.2014, DCF valuation was replaced by any internationally accepted pricing methodology on an arms’ length basis. Although DCF has become non-mandatory, it still remains the mainstay in all valuations for the purpose of FEMA compliance.

Section 56(2)(viib) of the income tax act – the infamous ‘angel tax’ tag is used for tax provisions in this section – is an anti-abuse provision that uses the fair market value of an unlisted company’s share as benchmark to compute a deemed income. Rule 11UA of IT Rules states that DCF value or Net Asset Value (NAV) – whichever is higher – should be taken as the fair market value for the purpose of Section 56(2)(viib). DCF value, being invariably higher than NAV value, is effectively only choice for computing the fair market value. Section 92 of the IT Act, however, does not prescribe any specific method for calculating the arms length price in related party transactions.

Given the universal preference for DCF among merchant bankers, valuers, and regulators, one tends to presume that DCF is a logical and mathematical method for arriving at the fair market value. Unfortunately, the method is pretty subjective and end result of valuation is highly sensitive to the valuer’s assumptions. It is the inherent flaws of this method that lead to litigation between a taxpayer and regulators.

DCF makes use of many assumptions and projections. The company undergoing valuation usually supplies its cash flow projections to the valuer. Where the company’s key assets are in the form of plant and machinery (such as manufacturing companies), the projections are quite accurate. Factors such as installed capacity, average capacity utilization, plant load factor, fixed and variable costs can be reasonably estimated. But when the company’s key assets are intangibles (IP, brands, trade secrets etc), the company can increase its revenue projections exponentially over the years. There is no limiting factor like in the case of manufacturing companies. Service sector companies are usually mired by optimism bias and do not take into account the possibility of redundancy and technological outage when making projections.

Cash flow projections can be the most arbitrary in the case of start-ups. Entire business of start-ups is based on an intangible asset called idea, which is the most troublesome intangible to quantify. A start-up that has just been incorporated and is not generating any revenue may make projections of astronomical revenue figures 5-6 years down the line. The founders are so confident of the idea that they started up; you cannot even blame them of any mens rhea in their optimism bias.

After the company supplies its cash flow projections, the valuer uses two key variables to find the enterprise value – discounting factor d and terminal growth rate g. If the valuer is not paid well, she assumes values of d and g based on ‘judgment’ and ‘prior experience with the sector’. If the valuer is paid well, she delves into the rigours of the Capital Asset Pricing Model (CAPM) to find d. This too is arbitrary, because the variables used in CAPM have to be assumed. Two variables in CAPM – beta value and firm-specific risk premium – are again to be determined by the valuer based on his best judgment.

As can be seen from above, there can be multiple differences of opinion between valuers when using a DCF-based method. Usually a small change in value of g or discount rate drastically changes the enterprise value.

Another way to highlight the sharp swings in DCF values due to incremental changes in discounting rate and terminal growth rate is to simulate a DCF using different discount rates and terminal growths. For this, I simulated DCF values of a real case having 5 years of cash projections using different discount rates and g. The results I got are as under (all enterprise value figures are in INR):


DCF Value Simulation





  factor (d)


Terminal growth rate (g)

















16. 0%








As seen from above, a change of terminal growth rate by 1%, on keeping d constant at 13%, leads to a change of 7.02% in EV from INR 31,69,605 to INR 33,92,250. Similarly, change in d by 2% (keeping g constant at 1%) leads to change of EV by -14.35%. If we change d from 13% to 17% and g from 1.5% to 0.5%, then the EV changes from INR 33,92,250 to INR 23,73,000. This is a change of 43% in EV!

A method that gives widely variant results on minor changes in assumptions and projections cannot be relied upon. High volatility increases risk of litigation under both FEMA and IT Act. This is a persistent problem with using the DCF method as a benchmark in tax and regulatory scrutiny.

The hullaballoo surrounding the so-called “angel tax” is precisely because DCF has failed start-ups. To a critic, future cash flow projections of start-ups – based on an idea their founders devoutly believe in – appear speculative. Small tweaks in discounting rate and terminal growth rate applied on these huge projections lead to high fluctuations in enterprise value.

Range of values as solution

In real life business dealings, a seller and an interested buyer – usually parties to a M&A process or investment round – meet and haggle over the price. They negotiate and price discovery happens at the end of the bargaining process. They use multiple methods (after judging the appropriateness of the methods) to arrive at a range of values. The prospective buyer starts negotiating from the lower end of the range and the seller starts negotiating from the upper end.

If the methods used are appropriate for the business, the range will be a narrow one. If the range one gets is wide, then a deeper investigation is required to find the reason for wide variance between fair values under different methods.

In any case, the Fair Market Value is a fiction. There is no real fair market value of unlisted securities; only a range of values within which the fair value lies.

In spite of its many deficiencies, DCF remains an appropriate method. However, it cannot give us a single fair market value. It can only give us a fair value range. There is, thus, a need for valuation practice to move from fair value calculations to fair price range calculation.

The writer has recently authored a book, “Tangible Guide to Intangibles: Identification, Valuation, Taxation, Transfer Pricing”. He is an officer of the Indian Revenue Service. Views expressed are personal.