How to value a bank stock in Pakistan, step by step.
A bank stock is valued from the balance sheet up: estimate deposits, split them into loans and investments, work out the spread, subtract costs and tax to get net profit and EPS, then apply a price-to-earnings multiple to reach a fair-value range.
- A bank is valued from the balance sheet up. Estimate deposits, split them into loans and investments, and work out the spread between what the bank earns and what it pays.
- That spread, minus the cost of running the bank and minus tax, becomes net profit. Divide by the share count and you have earnings per share (EPS).
- EPS turns into a fair-value estimate when you apply a price-to-earnings (P/E) multiple, then stress-test it with a sensitivity analysis.
- This is a framework, not a forecast. Every number below is rounded and illustrative; this site does not publish price targets or buy/sell calls.
Ask most people how they pick a bank stock and the answer is a dividend yield or a gut feeling. A bank is one of the few businesses you can actually model from the ground up, because its raw material is money, and money sits on a single page: the balance sheet. Estimate how much a bank will take in as deposits and what spread it will earn on that money, and you can build its profit, its earnings per share, and a defensible fair-value range, the price the shares would be worth if your assumptions hold.
This walkthrough lays out that method the way a brokerage analyst would, one line at a time. I'll use Meezan Bank (MEBL) as the reference point. It's Pakistan's largest Islamic bank and a stock plenty of retail investors already hold. One thing up front: every rupee figure below is deliberately rounded and illustrative, chosen to show how the arithmetic flows, not to forecast any company. This site does not publish price targets. The goal is that you can build and stress-test your own numbers.
01Why a bank is valued differently
For a normal company you forecast sales, subtract costs, and arrive at profit. A bank is different because both its costs and its revenue come from money. It pays depositors and lenders to raise funds (its cost of funds), then earns a return by lending that money out or investing it in government securities. The difference between the two is the net spread. At a conventional bank that's net interest income; at an Islamic bank like Meezan it's net profit on a profit-and-loss-sharing basis. Either way, it's the engine of the whole business.
So the lever that matters most is the cost of funding, and not all funding costs the same. This is where current accounts earn their reputation as a bank's most prized asset.
This is why analysts obsess over the CASA ratio, the share of cheap current and savings deposits in the funding mix. A bank that funds itself largely with zero-cost current accounts keeps far more of what it earns on its assets. When you read that a bank is targeting a "50% current-account share," that's management telling you they intend to protect the cheapest part of their funding.
02Step 1 - Write down your assumptions
A valuation is only as good as the assumptions under it, so the first move is to write them down. Spell them out and anyone reading your model (including you, six months later) can see exactly which lever to pull when reality turns out differently. For a Pakistani bank, a handful of assumptions drive almost the entire result.
| Assumption | What it controls | Illustrative value |
|---|---|---|
| Policy rate | The yield on assets and the cost of borrowings; the single biggest swing factor | 11.5% |
| Deposit growth | How fast the balance sheet, and therefore earning assets, expands | ~12% a year |
| Current-account share | The blended cost of funds (more current accounts = cheaper) | ~45% of deposits |
| Advances-to-deposit ratio (ADR) | How deposits split between loans and investments, plus the tax rate | ~50% |
| Tax rate | How much profit survives to shareholders | ~50% (ADR-linked) |
Notice the policy rate sits at the top. Because a bank earns on assets and pays on funding, a change in the State Bank's policy rate moves both sides of the spread at once. That's why a single rate decision can reshape an entire bank model. The SBP has held its rate at 11.5% through mid-2026 with CPI inflation around 7%, so the bigger question for any model right now is which way the next move goes. More on that in the risk section.
03Step 2 - Build the balance sheet
With assumptions in hand, grow the deposit base by your assumed rate, then decide what the bank does with that money. This is where the ADR comes in. It splits deposits into advances (loans to households and businesses) and investments (government securities, or for an Islamic bank, sovereign sukuk rather than conventional T-bills). Together, advances and investments make up the bank's earning assets: the pool of money on which it actually earns a return.
Two things deserve a pause here. First, money never sits idle. Whatever is not lent out is invested, so the bank earns on essentially its whole deposit base. Second, the ADR is not just an operational choice. In Pakistan it has historically carried a tax consequence, with banks below a threshold ADR taxed at a higher rate to nudge them toward lending. That's why the ratio shows up twice in our assumptions: once for revenue, once for tax.
04Step 3 - Build the P&L: the earnings engine
This is the heart of the model. Take the earning assets, apply a blended yield to get revenue, subtract the cost of funding and running the bank, subtract tax, and what survives is profit attributable to shareholders. Laid out as a ledger, each line follows from the one above it.
A few lines in that ladder are where bank models get their nuance:
- Cost of funds is not one number. It blends the profit paid on savings deposits (tied to the minimum deposit rate that regulation sets relative to the policy rate), the cost of any borrowings (close to the policy rate), and the zero cost of current accounts. Improve the current-account mix and this line shrinks.
- The repricing lag. When the policy rate jumps, a bank's assets don't all reprice instantly. Some loans are locked at older rates while funding costs rise immediately. Because assets and liabilities reprice on different schedules, a sharp rate move can briefly compress the spread until the book catches up. That's a normal asset-liability timing effect, not a one-off loss.
- Provisions set aside an estimate of loans that may not be repaid. For a well-managed bank in a stable year this is often a small line. It can spike in a downturn.
- Operating expenses grow with inflation and with branch expansion. A bank opening hundreds of new branches carries higher costs before those branches mature into deposits.
- The tax line is unusually heavy for Pakistani banks and, as noted, is linked to the ADR. That's why two banks with identical pre-tax profit can deliver different earnings to shareholders.
05Step 4 - Turn earnings into a value
Now you have an estimate of earnings per share. The final step is to decide what those earnings are worth. Two routes are common.
| Method | How it works | When it fits |
|---|---|---|
| Dividend yield | Value the share from the cash dividend it pays, relative to a target yield | High-payout stocks that distribute most of their profit |
| Price-to-earnings (P/E) | Apply an earnings multiple to estimated EPS to get a share price | Banks that retain a large share of earnings to grow (low payout) |
The choice hinges on the payout ratio, the slice of profit handed out as dividends. Take a bank earning around Rs 45 of EPS but paying only about Rs 18 a year in dividends, a payout ratio near 40%. Value that share on its dividend alone and you ignore the retained profit still building the business, so the dividend-yield method tends to understate a lower-payout bank. For that reason analysts usually lean on the P/E method for banks like this and treat dividend yield as a cross-check, not the main tool.
The P/E method is simple arithmetic: estimated EPS times a chosen earnings multiple. All the judgement sits in the multiple. A point either way changes the answer materially, so the honest move is to show a range rather than a single number.
What multiple is reasonable? That depends on the market's mood and the bank's quality. Pakistani banks have generally traded at modest single-digit-to-low-double-digit earnings multiples, partly a reflection of the macro risk premium investors attach to the market. With the KSE-100 around 181,000 in mid-2026, multiples have recovered from their worst but stayed well short of historic peaks. A stronger franchise tends to command the upper end of the range. A more uncertain backdrop, higher rates or geopolitical stress, compresses multiples across the board. The point of the sensitivity table is to make that judgement visible instead of burying it inside one number.
Trailing vs forward: don't mix them up. A stock screener shows a trailing P/E based on the last twelve months' reported earnings. The valuation above uses a forward P/E applied to estimated future earnings. If earnings are growing, the forward number looks different from the trailing one. They answer different questions, so never compare them directly.
06Step 5 - Total return, and the risks that move the answer
A share's payoff isn't just the price moving. It's the price change plus the dividends collected along the way. So the last piece is to add the expected dividend back to your price estimate for a total return. A bank paying a meaningful dividend can deliver a respectable total return even when the share price barely moves. For context, the larger PSX banks currently yield anywhere from MEBL's 5.5% up to MCB's 8.8% in our daily dataset, with UBL at 7.2% and HBL at 7.1% in between.
Every figure rests on the assumptions from Step 1, and some matter far more than others. Ranked roughly by impact:
- The policy rate, by far the biggest. Because it moves both the yield on assets and the cost of funds, a single SBP rate decision can rebuild the entire model. Stress-test this one first.
- Deposit growth. Faster growth expands earning assets and profit; a slowdown does the reverse. Model it conservatively to protect yourself from over-optimism.
- Current-account share. A few points either way shifts the cost of funds and the spread, though a well-run bank tends to keep this stable.
- The ADR. It affects both how much the bank lends and the tax rate it pays, so it touches the result twice.
- Operating costs and inflation. Usually a smaller swing, but worth flexing if inflation runs hot or branch expansion accelerates.
Run your model at a few different policy-rate and deposit-growth assumptions and you'll quickly see that a bank valuation is a range of plausible outcomes, not a single right answer. That humility is the most useful thing the exercise gives you.
Valuing a bank is a chain. Deposits, then earning assets, then spread, then net profit, then EPS, then a P/E-based fair value, then total return, with each link resting on a stated assumption. Build it yourself, stress-test the policy-rate and growth lines, and present a range rather than a point. The arithmetic is the easy part. The discipline is being honest about the assumptions. Nothing here is a recommendation to buy or sell any share, and the worked numbers are illustrative throughout.