Question:
For banks, is dividends the only practical way to estimate or value future cash flows?
Answer:
For bank equity, yes – in theory the value is the present value of all future distributions to shareholders – but that’s dividends + buybacks, and we often model it indirectly (ROE, P/B), not only as an explicit dividend DCF.
1. What is “cash flow to equity” for a bank?
For any company, equity value = PV of cash that eventually leaves the company and goes to you:
Dividends + share buybacks − new equity issued
For a bank, that’s basically it. There’s no clean “free cash flow to firm” because:
- Deposits are both “operating” and “financing”.
- Working capital swings are huge.
- Regulated capital (CET1, RWA) constrains what can be paid out.
So in practice, cash flow to equity for banks = dividends + buybacks, subject to capital rules.
Even when analysts don’t explicitly use a DDM, they’re implicitly valuing that stream.
2. How your justified P/B model fits in
Your formula
\text{Fair P/B} = \frac{\text{ROE} - g}{\text{CoE} - g}is just a repackaged dividend discount model / residual income model:
- ROE and g determine how fast book value grows.
- CoE is the discount rate.
- Payout ratio links earnings → dividends / retained earnings.
Mathematically, if you start from:
Value = PV of all future dividends
and assume constant ROE & g, you can rearrange the algebra and end up with your P/B formula. So even though dividends don’t appear in the final formula, they’re still in the background via:
- payout ratio,
- growth rate g.
So you’re already valuing dividend-like cash flows – just in a compact, book-value way.
3. Other ways people value banks (all equivalent in spirit)
Analysts typically use 3 families of models:
- Dividend Discount Model (DDM) / DCF to equity
- Explicit forecast of dividends (and sometimes buybacks) for 5–10 years.
- Then a terminal value (often via your same P/B logic).
- Painful but very “textbook”.
- Residual income / excess return (what you’re doing)
- Start from current book value.
- Add PV of future economic profits: (ROE − CoE) × equity.
- This collapses to the justified P/B formula when you assume constant ROE & g.
- Relative valuation (P/B, P/E, regression vs ROE)
- Look at how peers with similar ROE, growth, risk trade on P/B, P/E.
- Calibrate your target multiples using a simple DDM or justified P/B in the background.
Even when people talk only about P/E or P/B, if you push them, the story always comes back to:
“Because I think they can earn X% ROE, grow Y%, and pay out Z% of earnings over time.”
Which is just a dividend / cash-flow story in disguise.
4. What about banks that never pay dividends (like PNLF)?
Three possibilities:
- Temporary 0% payout (building capital, early growth phase)
- You still assume a positive long-term payout in the model.
- DDM / justified P/B works fine with a separate “high-retention” stage then “normal” stage.
- Perpetual hoarder but ROE > CoE (rare)
- Theoretically, if they reinvest at high ROE forever, your wealth grows via BVPS compounding.
- Eventually someone will demand a dividend / buyback or buy the bank.
- Perpetual hoarder with ROE < CoE (PNLF-type story)
- Every rupiah retained earns less than your required return.
- In theory, fair P/B can be far below 1x; low PBV is justified.
- The problem becomes governance, not finance: will that trapped value ever be released?
In cases like (3), your justified P/B framework is very powerful: it explicitly tells you “low PBV is not necessarily cheap if ROE is structurally bad and cash never comes back.”
5. Takeaways for your mental model
- For bank equity, everything ultimately comes back to dividends & buybacks, even if you model via ROE & P/B.
- Your justified P/B formula is not an alternative to cash-flow valuation – it is a compressed form of it.
- “Cheap on PBV” is only attractive if:
- ROE ≥ CoE (or can get there), and
- Management will eventually return excess capital (reasonable payout / buybacks).
