Statistical Valuation


Valuation is an art of arriving at an optimal value of a company. Various methodologies are being used for the same but the fundamentals are based on the expected growth of the sector and the cost of capital. This works well when one uses this methodology on companies within a given sector. This is an easy part. A complication arises when a given company has expanded into multiple businesses and more importantly into areas which have no relation to one another.

For all the time and money companies invest in integrating separate businesses into a single strategy and culture, most managers also understand that different businesses have their own management challenges. Although they see their companies as more than the sum of their parts, they also create separate management, P&L, strategic plans, and performance targets for each of their business units. This gives freedom to take independent decisions irrespective of the functioning of other areas. But when it comes to valuing the company these same managers tend to use a single cost-of-capital matrix or valuation multiple across the entire organization instead of breaking these measures out business by business, the way many outside observers and securities analysts do.

For companies with distinctly different businesses, it doesn’t make sense to use the same cost-of-capital assumptions or valuation multiples or even an average of them when assessing these different businesses’ performance or strategic plans.


Multi Business Valuation

Multi-business valuation is useful for determining whether a company is more valuable as a combination of businesses or whether it is more valuable if the units are operated as stand-alone entities. It helps to create a clearer picture of headquarters costs and benefits since headquarters can be valued as if it were a separate cost center.

Valuing a multi-business company is fundamentally the same as valuing a single-business company. What makes multi-business valuation more complex is that each business unit has its own cash flows, capital structure, and cost of capital. In addition, business units may have shared cash flows and it may be hard to separate the costs and benefits of corporate headquarters. Valuing a multi-business company is somewhat like putting together building blocks. The value of the entire corporation is the sum of the value of the business units, plus non-operating assets, less the unallocated costs of corporate headquarters. Unique issues in valuing a multi-business company that was not discussed earlier in this book include:

  • Defining business units and their cash flow.
  • Estimating business-unit capital structure and cost of capital.


Defining Business Units and Their Cash Flows

In principle, a distinct business unit could be split off as a stand-alone business or sold to another company. A good rule of thumb is to define business units at the smallest practical level of aggregation.

To identify business-unit cash flows, one has to deal with two typical problems: transfer pricing and corporate overhead. Transfer pricing arises when the output of one business unit is the input of another. A high price will increase profits of the supplier at the expense of the user, and vice versa. The recommended solution is to establish a transfer price as close as possible to the buttonet price of close substitutes. This can be a difficult task if substitutes are hard to find. The main idea is to approximate buttonet prices so that profit is appropriately allocated to the supplier or to the user of the goods or service.

Taxation often complicates transfer pricing. One of the benefits of the corporate umbrella (that is, of headquarters) is that it can establish a transfer-pricing system that keeps profits in the jurisdiction that has the lowest tax burden. As a result, one set of artificial transfer prices may be used for tax purposes and another set of buttonet-determined prices for determining business-unit cash flows (although tax authorities may not permit this).

Allocating corporate overhead is a problem closely related to transfer pricing. The central issue is whether business units would use corporate services if they were spun off as separate entities. Many services would be used; for example, accounting, legal, computer, and internal consulting services.

Whenever possible, the cost of these services should be allocated to business units on a usage basis or, failing that, on the basis of a reasonable proxy for usages, such as operating income, revenues, capital employed, or a number of employees. In theory, buttonet prices should be used for allocation purposes.


Valuation of Multiple Business-2

Estimating Business-Unit Capital Structure and Cost of Capital

The capital structure of each business unit should be consistent with that of comparable companies in its industry and with the overall philosophy of its parent. If the parent is aggressive and chooses a particular bond rating, then business units will normally have capital structures that bring them the same rating within their industry. At that rating, different business units may be able to borrow varying amounts of debt. An insurance subsidiary may be able to carry an 80 percent debt-to-capital ratio, while a manufacturing subsidiary might only have a 25 percent debt-to-capital ratio. The basis for determining business-unit capital structure will usually depend on cash flow or capital employed at the business unit. Debt-to-capital ratios or interest-coverage ratios for comparable companies are a useful starting point.

The difference between the sum of all business-unit debt and total companywide debt should be attributed to corporate headquarters. Any remaining debt results from the fact that the debt capacity of a portfolio of business-unit cash flows that are not perfectly correlated has less variance than the sum of the separate cash flows. Consequently, the combination of separate business units under a corporate umbrella provides greater debt capacity. The present value of the interest tax shield is a benefit of headquarters.

Having determined the target capital structure and a tax rate of each business unit, one still needs to estimate the cost of equity to establish its weighted average cost of capital. In the case of a division of a company or a nonpublic company, no betas are published. To estimate a beta, one of the three approaches is used: management comparisons; comparison companies, or a multiple regression.

  • Management Comparisons :One of the methods revolves around getting the management to give their views on the risk-to-reward ratio of each of the business unit. This method generally works under most conditions as the managers of the company have an insight view of the proceedings of affairs of the company and the sector to which their business belongs at large.
  • Comparison Companies :A second approach is to identify the publicly traded competitors most similar to the division. One can then look up the betas for these companies, which are presumed to have a similar risk. But there is a catch. Beta is a measure of the systematic risk of the levered equity of the comparison companies, and these companies may employ leverage differently from that used by the division one is attempting to value. To get around this problem, one has to un-lever the betas of the comparison companies to obtain their business risk, then re-lever using the target capital structure of the division one is analyzing.The unlevered beta measures the business risk of a company by removing the effect of financial leverage. The observed equity beta computed from buttonet return data presents a picture of the risk of equity given the company’s existing leverage. To un lever the beta, one needs data on the company’s levered beta, its target capital structure, and its marginal tax rate.
  • Multiple Regression Approach :While using one of the traditional methods of valuation i.e. Relative Valuation, it is necessary to first understand target’s business model. It becomes important to break the structure into parts which are aligned under one roof.One obstacle to breaking these measures out is that the traditional approach requires managers to compare their businesses with others to derive a benchbutton. Often, so-called pure-play comparisons with similar characteristics and performance may not exist. And even where there may seem to be an abundance of comparable businesses, identifying the right ones is often as much a matter of personal pride as of getting a hold of enough data. A way around this problem is to recognize that the business risk (that is, the unlevered beta) of a multidivisional company is a weighted average of the risks of each line of business.For many companies, statistical modeling or more precisely, regression analysis can offer an alternative. It can help them more accurately estimate business unit level betas and valuation multiples even where there are few pure-play comparisons, enabling them to make better trade-offs between businesses. Such modeling works in industries where many companies have different mixes of similar businesses such as banking, chemicals, insurance, metals and mining, and technology.In models used by Mckinsey & Company, whereby they start by assessing the business mix by revenue or asset weights the independent variables and regressed those weights against company-level multiples or betas the dependent variables. The regression model then determines the betas or multiples for individual business units. With this approach, managers can draw on a broader range of companies even including those they would not normally see as peers and let the statistical regression control for differences in the size and makeup of their portfolios. As a result, analysts are more likely to find sufficient available data.When changes in risk are expected, be prepared to re-estimate an equity beta. Sometimes a company’s strategic plan implies that risk is expected to change. A young company, recently gone public, is risky now (high beta), but is expected to have declining risk across time. Although one cannot suggest foolproof steps for estimating changing risk, it will change. This implies that the weighted average cost of capital may also change (decline) as the company matures. Consequently, it becomes necessary to discount year-N cash flows at a risk-adjusted rate approximate for the risk in year N, not other years when the risk might be higher or lower.


See some more related Articles on Valuation.

Case Study

To understand the case better let’s take a case of a well-known conglomerate based out of India. The company is in multiple businesses which include Retail, IT, Telecom, Financial Services and Engineering Goods. Looking from the perspective of valuation we can break the company into different business units as if they were independently functional companies.

Business Shareholding (%)
Retail 100
IT 100
Telecom 25
Financial Services 75
Engineering Goods 100

Business Cash Operating Profit(INR Cr.)
Retail 200
IT 200
Telecom 5000
Financial Services 600
Engineering Goods 800

Business Debt-Equity Ratio
Retail 0.8
IT 0.01
Telecom 1.00
Financial Services 4.45
Engineering Goods 1.3

Business EV-EBIDTA (X)
Retail 20
IT 14
Telecom 8.5
Financial Services 4.3
Engineering Goods 4

Business EV(INR Cr.)
Retail 4,000
IT 2,800
Telecom 42,500
Financial Services 2,80
Engineering Goods 3,200

Business EV as per Shareholding(INR Cr.)
Retail 4,000
IT 1,628
Telecom 10,625
Financial Services 1,935
Engineering Goods 3,200

Net Enterprise Value of the company – INR 21, 400 Cr.



Valuing a multi-business company by components often leads to a critical rethinking about what business units really belong as part of the company. Frequently managers will decide to restructure the company and focus on a smaller scope of activities. Managers are then faced with how to dispose of units that no longer fit. Over the past decade or so, the choice of restructuring techniques has widened as different techniques have been perfected and accepted by the stock buttonets. In addition to selling a business unit to another company, managers can now choose among spin-offs, equity carve-outs, track stock, and management buyouts. Valuation is a key tool for managers to use to determine what is best for their shareholders among all the options.

This article has appeared first in magazine M & A Critique in January 2013.

Main Contributor : Shivraj Singh

If you like the article share it with your colleagues.

Post a comment