DCF (Advanced) Flashcards
Explain why we use the mid-year convention in a DCF.
- You use it to represent the fact that a company’s cash flow does not arrive 100% at the end of each year - instead, it comes in evenly throughout each year.
- In a DCF w/o the mid-year convention, we would use discount period numbers of 1 for the 1st Year, 2 for the 2nd Year, 3 for the 3rd Year and so on.
- WITH the mid-year convention, we would instead use 0.5 for the 1st Year, 1.5 for the 2nd Year, 2.5 for the 3rd Year and so on.
- The end result is that the mid-year convention produces higher values since the discount periods are lower.
What’s the point of a “stub period” in a DCF? Can you give an example?
- You use a stub period when you’re valuing a company before or after the end of its fiscal year and there are 1 or more quarters in between the current date and the end of the fiscal year.
- For example, it’s currently Sep. 30 and the company’s fiscal year ends on Dec. 31.
- In this case, it wouldn’t be correct to assume that FCF only starts on Jan. 1 of the next year - there are still 3 months between now and the end of the year, the company still generates FCF in those 3 months, and you need to account for it somewhere in your model.
- So you would calculate FCF in that 3-month period, use 0.25 for the discount period, and then use 1.25 for the discount period for the first full year of the model, 2.25 for the next year, and so on.
What discount period numbers would you use for the mid-year convention if you had a stub period - e.g. Q4 of Year 1 - in a DCF?
- The rule is that you divide the stub discount period by 2, and then you simply subtract 0.5 from the “normal” discount periods of the future years. (See p. 53 for example of a Q4 stub).
- See (p. 53-54) another example for a Q2-Q4 stub (e.g. we’re valuing the company on Mar. 31 and its Fiscal Year ends on Dec. 31)
- What is the logic here? Think about it like this: let’s take the example of the normal discount period for Year 1 being 1.75, representing 3 quarters and then a full year.
- Now ask yourself when you receive that cash flow in Year 1. You’re still receiving it midway through that 1st Year, in other words, you still use 0.5 of that period.
- However, you also need to take into account the 3/4 of this partial year b/c 3 quarters pass between now and the start of Year 1. So you still have 0.75 there, and the mid-year discount period with the stub period is 0.75 + 0.5, or 1.25.
- That is why it’s not 1.75/2 like you might expect: it’s about when you receive that cash flow in a given Year, from the perspective of the start of Year 1 - and then you add the total amount of time that passes between now and the start of Year 1. There’s no mid-year discount applied there b/c we don’t receive any Year 1 cash in this 1st partial year.
How does the Terminal Value calculation change when we use the mid-year convention?
When you’re discounting the Terminal Value back to its present value, you use different numbers for the discount period depending on whether you’re using the Multiples Method or Gordon Growth Method:
• Multiples Method: You add 0.5 to the final year discount number to reflect that you’re assuming the company gets sold at the end of the year.
• Gordon Growth Method: You use the final year discount number as is, b/c you’re assuming the FCFs grow into perpetuity and that they are still received throughout the year rather than just at the end.
What if you have a stub period AND you’re using a mid-year convention - how does Terminal Value change?
It’s the same as what’s described previously - a stub period in the beginning does not make a difference.
How does a DCF for private company differ?
- The mechanics are the same, but calculating Cost of Equity and WACC is problematic b/c you can’t find the market value of Equity or Beta for private companies.
- So you might estimate WACC based on the median WACC of its Public Comps, and do the same for Cost of Equity if you’re using that as the Discount Rate.
How do you factor in one-time events such as raising Debt, completing acquisitions, and so on in a DCF?
- Normally, you ignore these types of events b/c the whole point of calculating FCF is to determine the company’s cash flow on a recurring, predictable basis.
- If you know for a fact that something is going to occur in the near future, then you could factor that in - issuing Debt or Equity would change Cost of Equity and WACC (and the company’s FCF in a levered DCF); completing an acquisition or buying an asset would reduce cash flow initially but perhaps boost it later on.
What should you do if you don’t believe management’s projections in a DCF model?
You can take a few different approaches:
• You could create your own projections.
• You could “hair-cut” management’s projections (reduce them by a certain percentage) to make them more conservative.
• You could show a sensitivity table based on different growth rates and margins, and show the values using both management’s projections and a more conservative set of numbers.
Why would you not use a DCF for a bank or other financial institution?
- Banks use Debt differently than other companies and do not use it to finance their operations - they use it to create their “products” - loans - instead.
- Also, interest is a critical part of banks’ business models and changes in “Operating Assets and Liabilities” can be much larger than a bank’s Net Income. Finally, CapEx does NOT correspond to reinvestment in business for a bank, and is often negligible.
- For financial institutions (commercial banks and insurance firms), it’s more common to use a Dividend Discount Model or Residual Income Model instead of a DCF.
Walk me through a Dividend Discount Model (DDM) that you would use in place of a normal DCF for financial institutions.
The mechanics are the same as a DCF, but we use Dividends rather than Free Cash Flows:
1. Project the company’s earnings, down to Earnings per Share (EPS).
2. Assume a Dividend Payout Ratio - what percentage of the EPS gets paid out to shareholders in the form of Dividends - based on what the firm has done historically and how much regulatory capital it needs.
3. Use this to calculate Dividends over the next 5-10 years.
4. Do a check to make sure that the firm still meets its required Tier 1 Capital Ratio and other capital ratios - if not, reduce Dividends.
5. Discount the Dividends in each year to their present value based on Cost of Equity - not WACC - and then sum these up.
6. Calculate Terminal Value based on P/BV and Book Value in the final year, and then discount this to its present value based on the Cost of Equity.
7. Sum the present value of the Terminal Value and the present values of the Dividends to calculate the company’s net present value per share.
• The key difference compared to a DDM for normal companies is the presence of capital ratios - you can’t just blindly make Dividends per Share a percentage of EPS.
Do you think a DCF would work well for an oil & gas company?
If it’s an Exploration & Production (E&P)-focused company, generally a DCF will not work well b/c:
1. CapEx needs are enormous and will push FCF down to very low levels.
2. Commodity prices are cyclical and both revenue and FCF are difficult to project.
• For other types of energy companies - services-based or downstream companies that just refine and market oil and gas - a DCF might be more appropriate.
• For more on this topic and the alternative to a DCF that you use for oil & gas companies (called a Net Asset Value, NAV, analysis) see the industry-specific guides.
How does a DCF change if you’re valuing a company in an emerging market?
- The main difference is that you’ll use a much higher Discount Rate, and you may not even necessarily link it to WACC or Cost of Equity, because there may not even be a good set of Public Comps in the country.
- You might also add in a premium for political risk and uncertainty, and you might severely reduce management’s growth or profit expectations, especially if they have a reputation for being overly optimistic.
When you’re calculating WACC, do you count Convertible Bonds as real Debt?
- Trick question. If the Convertible Bonds are in-the-money then you do not count them as Debt, but instead assume that they contribute to dilution, so the company’s Equity Value is higher.
- If they’re out-of-the-money, then you count them as Debt and use the interest rate on the Convertible Bonds for the Cost of Debt (and include them in Debt in the formula for Levered Beta).
What about the treatment of other securities, like Mezzanine and other Debt variations?
- If interest is tax-deductible, you count them as Debt in the Levered Beta calculation; otherwise, they count as Equity, just like Preferred Stock.
- For WACC itself, you normally look at each type of Debt separately and assume that the “Cost” is the weighted average effective interest rate on that Debt.
Should you ever factor in off-Balance Sheet Assets and Liabilities in a DCF?
- Potentially, yes, especially if they have a big impact on Enterprise Value and Equity Value (i.e. if they’re something that the acquirer would have to repay).
- But it’s not terribly common to see them, partially b/c when off-B/S items are more important (for commercial banks w/ derivative books, for example), you don’t even use DCF.
