Notes for Technical Interview Questions

During a recent round of interviews with several banks, headhunters sent across a number of potential technical questions that they said I should be very comfortable with. However, through all of my interviews I was never actually asked any technical questions (although I did do a 3 hour modelling exam for one bank). Perhaps this is because I was coming from a couple of years already in IB and the interviewers assumed I knew what I was doing. Still, I took the time to put my thoughts in order. If any of you would like a refresher, you can find some of my notes below.

Just a small point to start with – Valuation is often an art and not an exact science. There are many different ways to approach the numbers, often depending on what you (and your client) want the numbers to show.

It is critical, however, that you can stand in front of a group of senior Japanese executives, for example, and confidently explain to them why your numbers stack up the particular way they do. Similarly, if you are preparing for an interview by simply memorizing the below formulas, but you cannot smoothly explain the function of each piece of each formula, you will not (or at least you should not) proceed to the next round.

How would you value a company?

- Depends on the industry and situation of the company

- We usually use a combination of tools, namely:

o Discounting the future earnings to the present value, which would generally be in the form of a DCFF, DCFE or DDM, depending on the company’s industry and individual situation

o Using a multiple of earnings like EV/EBITDA or P/E, or a multiple of the company’s assets like oil reserves or book value, depending on the industry and situation

How do you do a DCFF?

- First model out the future earnings of the company, ideally with the help of management estimates, broker estimates, maybe some third party consultant figures, and your own best judgements

- After you get the forecast, you find the Free Cash Flow to Firm in each year:
+ EBIT
- Tax on EBIT
- Capex
+ Depreciation + Amortization
- Increase in WC assets
+ Increase in WC liabilities
+ Any other cash or non-cash adjustments that are company specific
= FCFF

- For the terminal value, at the end of the forecast period, you would use either:

(1) The Gordon Growth Model: (Final Year FCFF * (1 + Perpetual Growth Rate) ) / (WACC - Perpetual Growth Rate)

(2) Exit Multiple, which could be based on the entry multiple, or the long term average multiple for the industry, depending on the situation

- Then you would discount the FCFFs to the present value
o For this purpose, you would combine the Terminal Value with the annual free cash flow in the final forecast year
o FCFF in a particular year / (1+ WACC) ^ number of years in the future that particular cash flow occurs

- This would give you the Enterprise Value

o To get the equity value, you would:
+ Enterprise Value
- Minority Interests
- Net Debt
- Unfunded Pension Liabilities
- Preferred Shares
+ Associates / JVs
= equity value

How do you derive FCFE from FCFF?

o I usually start from EBIT when I am doing a DCFF, but I usually start from Net Income when I do a DCFE
+ Net Income
+ Depreciation & Amortization
- Increase in WC assets
+ Increase in WC liabilities
+ Increase in Deferred Tax
- Capex
- Interest on cash (net of tax)
+ Drawdown/(Repayment) of debt
+ Any other cash or non-cash adjustments that are company specific
= FCFE

o If you for whatever reason had to start with FCFF:
+ FCFF
+ Tax on EBIT
- Actual tax paid
- Interest on cash (net of tax)
- Interest on debt
+ Drawdown/(Repayment) of debt
= FCFE

What is the discount rate used?

- The discount rate reflects the cost of capital that correlates to the cash flows you are discounting

- If you are doing a DCFF, then you would use a WACC, since it accounts for both Debt and Equity capital and the cash flows you are discounting are "pre-financing" and do not already include interest expense

- If you are doing a DCFE or a DDM, then you would use just the Cost of Equity since the cost of debt has already been taken into account in the cash flows that you are discounting

How do you get WACC? What is the tax rate used - marginal or effective?

+ Cost of Debt:

o Debt / (Debt + Equity)
- Should generally be the target capital structure, however we often use a rolling WACC, which changes with the capital structure of each forecast year

o * (Cost of Debt)
- Usually just the interest expense over the principal outstanding)

o * (1 – tax rate)
- This should generally be the marginal tax rate. We usually refer to a KPMG guide for this. However, for some companies that operate in a wide variety of tax jurisdictions, it would be incorrect to just use the marginal tax rate in Singapore, for example, for their global operations, in which case the effective tax rate could sometimes serve as a proxy. Similarly, if for some reason the company expects to continue to enjoy some kind of tax breaks, such as those enjoyed by some foreign invested, pure-export companies in China, then we might be better off using an effective tax rate. Usually we would just use the marginal rate though

+ Cost of Equity

o Equity / (Debt + Equity)
- Should generally be the target capital structure, however we often use a rolling WACC, which changes with the capital structure of each forecast year

o * Cost of Equity
- Usually based on CAPM, although some clients have a specific cost of equity that they like to use

What is the cost of debt and cost of equity used in calculation of WACC? How would you find them?

- Cost of Debt

o We usually take the effective interest being paid on the book value of the principal outstanding

o This should generally reflect the true cost of all of the company’s debt, however if the company is in a distressed situation, you may be better to revalue the debt.

- Cost of Equity

o Use the CAPM formula:
- Risk Free Rate + Beta * (Market Return – Risk Free Rate)

o Risk Free Rate
- Generally we would use the long term local currency government bonds if, for example a Singaporean company was going to invest in a domestic project in Indonesia
- If you are acquiring a diversified international company like an international trading company based in Singapore, you could use the Singapore govt bonds.

o Beta
- If this is for a domestic Indonesian project, you would ideally look at the betas of comparable listed companies in Jakarta vs the Jakarta index. You would unlever the betas, generally take the average and then relever the beta based on the project capital structure
- If this is for an acquisition of an international company based in Singapore, then you would do the same with comparable companies listed on the SGX
- (if you wanted to use a rolling cost of equity, you would relever each year based on the changing projected capital structure)

o Market Return
- Similarly with the Risk Free Rate, if this is a Singaporean company investing in a domestic Indonesian project, you would use the market return on the Jakarta stock exchange
- If you are looking at a Singaporean company, you would look at the market return on the SGX
- Damodaran has a list of market returns for different countries, but we would also sometimes ask our country team for their general practice, as they are supposed to have a deeper understanding of the local market
- If you are arriving at a result that you are not happy with, this market return provides a convenient adjustment mechanism, since it is a relatively soft figure

How do you get the terminal value and how would you get the main assumptions?

- Either:

(1) Gordon Growth Model:
o (Terminal Year Cash Flow * (1 + terminal growth) ) / (Discount Rate – terminal growth Rate)

- Terminal Year Cash Flow
o Would be either FCFF, FCFE, or Dividend depending on what valuation you are doing

- terminal growth
o Usually a very conservative assumption, a bit higher than forecasted long term nominal economic growth rate

- Discount Rate
o Would be either WACC or Cost of Equity depending on valuation method

(2) Terminal multiple
o EV/EBITDA, P/E or other relevant multiple

- Could be based on the entry multiple, or the long term average multiple for the industry, depending on the situation

- This is sometimes preferable because the GGM is so sensitive to unknown assumptions, while an exit multiple is more understandable

Case study of what are the modelling steps for modelling a merger. How do you get goodwill?

- Make forecasts for each individual company

- Balance Sheet

o Determine:
- The purchase price of the acquisition
- Funding for the acquisition
- Transaction costs (such as advisory fees)

o Then make a sources and uses (this is helpful to determine how much debt would have to be raised)
- Sources
• Existing cash
• New Debt
• New Equity
- Uses
• Acquisition price (May include buying out stock options)
• May include paying back target company debt holders, although not always
• Transactions costs (like advisory fees)

o Determine Good Will
+ Acquisition cost
– Net asset value of target, which usually equals the shareholder equity value excluding minority interests
= Good Will

o Adjust parent balance sheet:
- If the parent is issuing new debt or equity, you would put those transactions through (effects cash as well as debt or equity)
- If the parent is paying cash, then the cash would come out of the parent’s balance sheet (effects cash)
- If the parent is paying in new shares, then we’d reflect the increase in parent share capital (effects equity)
- Expense the transaction fees (effects cash and retained earnings)
- Add the Good Will from the transaction (effects Assets)

o Consolidate parent and target
- Remove all shareholders equity from target (but keep minority interests)
- Then consolidate adjusted parent with adjusted target

- Income Statement

o Add in any synergies (such as increased revenues, increased cost efficiencies) to parent and/or target
o Add any expenses related to the issuing of new debt
o Subtract any tax charges arising from any additional profit
o Something to note: While the parent company would not consolidate the retained earnings of the target at the time of acquisition, the parent company would generate new retained earnings on future net income of the target

Bring me through the main differences between transaction multiples and trading multiples

- Transaction multiples theoretically include a take-over premium
o However, in some industries, there is not a depth of comparable and recent transactions
o Furthermore, the financials for many private transactions are based on press reports and other unofficial non-transparent sources

- We generally rely more on trading multiples because the data is more accurate, more recent, and perhaps more commercially relevant for the target company valuation

- We would then generally add an overlay indication with a takeover premium of maybe 25% or so, but this is a sliding number based on the unique situation

- We would also do a rolling multiples over time, to determine whether current valuations are depressed and some shareholders may be expecting to exit at a multiple closer to ‘the good times’

How do you get Enterprise Value - detailed to the minority interest, associates etc.

Well, if you are starting with equity value:
+ equity value
+ Net Debt
+ Preferred Equity (although it’s not common in Asia)
+ Unfunded Pension Liabilities (although it’s not common in Asia)
+ Minority Interests
– Associates/JVs
= Enterprise Value

How would you do a merger by way of issuance of shares? What is the difference between the issuance of shares and cash?

- Build a Sources and Uses
- Instead of using current cash and new debt as sources, you would use new equity as a source
- You would then determine at what value you would be issuing each new share (Maybe a slight discount to current share price)
- You could then determine how many new shares you would need to issue
- Break that into the par value and additional paid in capital
- This is of course going to dilute your current shareholders, but if the current shares are trading at a high valuation, shares could be a cheap currency to use
- Of course you would want to check how this whole process will affect the earnings per share for your current shareholders

How would you know if an acquisition is dilutive?

- If your current shareholders’ earnings per share goes down after the transaction, this would be dilutive
- If your current shareholders’ earnings per share goes up, then it would be Accretive

- It is best to look at the effects over a number of years, otherwise, this could be a bit short sighted
- For example, if a pharmaceutical company with a low P/E acquires a small company with a new miracle drug, but that new miracle drug may not be on the shelves yet, the transaction is probably going to be dilutive for the buyer’s shareholders. However, this new miracle drug might prove to be very lucrative, especially when combined with the buyer’s existing distribution network. So in the medium term, EPS might improve and investors, who are looking at future earnings, might even push up the stock price despite the transaction being immediately dilutive

How do you calculate beta?

- Plot the index in one column and the relevant stock price in the next column (this might be closing data for each week over the past 3 years, for example)
- In the next column, calculate the % change in the index each week, and in the next column, calculate the % change in the stock price each week
- You would then take (COVAR(column containing % changes of the stock price, column containing % changes of the index) / (VAR(column containing % changes of the index)
- You would then delever the beta to get a beta that you can relever in subsequent calculations

What is the difference between asset beta and equity beta?

- The asset beta, which I would usually call the unlevered beta is the risk of the asset, without the additional risk of any leverage that the asset has
- Because more leverage means more risk for the equity holder, when you calculate the beta of any stock, this beta naturally takes into account this leverage
- When you are trying to figure out a beta for your target company by looking at the betas of the comps, you would calculate each comp’s levered (or equity) beta, then delever it
- You would then re-lever the beta based on the capital structure of your company

How would you get a beta for a company which is private?

- Come up with a list of comparable relevant companies
- Calculate the equity (or levered) betas for each of these companies, relative to the relevant index
- Delever each of the betas based on each company’s capital structure
- You would then be able to calculate the average unlevered beta for the group
- You can then take that unlevered beta and reliever it based on your company’s capital structure

Note to those who have followed some of my previous postings:

I was in IB for a couple of years. I then I took a year off to try to build a tech startup. I have now returned to IB.

I am planning to put together a thoughtful piece reflecting on this transition, however I will not be writing this until January or February in order to more clearly and holistically present the lessons learned.

In the meantime, feel free to comment on any of the technical points above.