A governance view has to arrive in the model through exactly one of two doors, and it must never come through both.
The first door is the discount rate. Add basis points to the cost of equity to reflect the probability and size of value diverted from the minority, and every future cash flow is worth less. The second door is a multiple haircut: leave the discount rate alone and pay a lower multiple of the same earnings. In a constant-growth model the two are the same statement. If value is FCFE / (Re - g), then any add-on to Re maps to a unique proportional cut in the multiple, and quoting both is double-counting.
How big should a governance adjustment be?
Watch the size. With a cost of equity of 13.00% and growth of 6.00%, one point of governance add-on cuts the value multiple from 14.285714 to 12.500000, a fall of 12.5000%. Hold the same add-on and drop growth to 3.00% and the fall is only 9.0909%. The same governance opinion is worth more in a faster-growing company, which is not obvious and which is a good reason to state the adjustment in basis points and let the model translate it.
The size has to be defended, and there is no measurement that will do the defending for you. What works is a stated, capped rule, applied identically to every company you cover, whose maximum effect you can name. The rule used throughout what follows is a governance add-on equal to a 3.00 percentage-point cap multiplied by the fraction of the checklist a company fails.
That is a house convention and nothing more. Its virtue is that the cap is explicit, so the worst governance opinion available can take at most 27.7136% off Vindhya Ferro's equity value, and a reader who thinks that is too much or too little can move one number and watch the whole answer move.
Advertisement
How do you keep the adjustment honest?
Two disciplines keep the adjustment honest. Make it separable, meaning a single cell that can be set to zero to recover the unadjusted valuation. And make it monotone in evidence, meaning each checklist row moves the answer by a stated amount, so that fixing the row is worth a stated sum. At Vindhya, the chair-separation row and two independent-director appointments together lift the score from 6 to 10 out of 20 and are worth ₹136.1417 crore of equity value, or ₹34.0354 crore per point of score.
A board that will not pay a recruiter ₹2 crore to hire two independent directors is refusing a trade at better than sixty to one on the analyst's own arithmetic. That is a sentence worth writing in a note.
What does the research say about governance and returns?
Gompers, Ishii and Metrick built the best-known attempt at the empirical version, a governance index over twenty-four provisions, and reported a return gap between the strongest-rights and weakest-rights portfolios across the 1990s (Quarterly Journal of Economics, 2003). Later work has argued about whether the effect survives out of sample and about what drives it. Treat the literature as a reason to believe the sign and not as a source for the size, and do the sizing yourself, in the open.