CFA Exam Calculator Tips: BA II Plus Shortcuts

Nobody fails a CFA exam because their calculator was slow. People lose marks because they re-keyed a number and fat-fingered it, because P/Y was still 12 from the previous item, or because they computed the population standard deviation when the question wanted the sample one.

All three are fixable in an afternoon, and none of them is about knowing more finance. This is the operational layer — the part that separates candidates who finish with fifteen minutes spare from candidates who guess on the last six items.

Ninety seconds an item

Level I gives you roughly ninety seconds per multiple-choice item, averaged across the whole session. Quantitative items are not evenly distributed through that budget: reading and setting up a time-value-of-money question takes most of the ninety seconds, and the keystrokes take eight.

So the arithmetic is not the constraint. What is:

• Transcription. Writing an intermediate result on paper and typing it back in. Two chances to make an error, and about six seconds each way.
• Leftover settings. A wrong P/Y, a stale FV, BGN still on. These do not produce error messages, they produce plausible numbers.
• Hesitation. Not being certain whether the question wants Sx or σx, or whether the payment is at the beginning or end of the period, and re-reading the stem twice to decide.

Everything below attacks one of those three.

What to compute in a worksheet rather than a formula

The formula sheet is a fallback, not a method. Most quantitative items map to a worksheet that gives the answer directly, and the mapping is worth memorising because it removes the hesitation entirely.

The stem says	Use	Not
“effective annual rate, given a nominal rate”	2ND ICONV	(1 + i/m)^m − 1 by hand
“sample standard deviation of these returns”	2ND DATA then 2ND STAT	the summation formula
“monthly rate equivalent to an effective annual rate”	y^x with 1 ÷ 12	dividing by 12
“net present value of these cash flows”	CF then NPV	discounting each flow separately
“internal rate of return”	CF then IRR	trial and error
“price of this bond”	2ND BOND, or the TVM row on a coupon date	manual discounting
“interest paid in year three of the loan”	2ND AMORT with P1 = 25, P2 = 36	building a schedule
“value of a perpetuity”	plain division: 5 ÷ 0.065 = 76.9231	forcing a large N into TVM
“days between two dates”	2ND DATE	counting months

The perpetuity row is the one candidates over-engineer. A level payment forever is a division, and keying 9999 N into the TVM row to approximate it wastes fifteen seconds to arrive at the same number.

Sx or σx: the trap that changes the answer choice

The 2ND DATA worksheet takes raw numbers; 2ND STAT reads the summary statistics out of them. Two of those are standard deviations, and choosing wrongly is the single most common quantitative error on Level I.

Six annual returns on a fund: 4.2%, −1.8%, 7.6%, 2.9%, −3.4%, 9.1%.

1. 2ND
2. · DATA
3. · 2ND
4. · CLR WORK
5. · 4.2
6. · ENTER
7. · ↓
8. · ↓
9. · 1.8
10. · +|−
11. · ENTER
12. · ↓
13. · ↓
14. · 7.6
15. · ENTER
16. · ↓
17. · ↓
18. · 2.9
19. · ENTER
20. · ↓
21. · ↓
22. · 3.4
23. · +|−
24. · ENTER
25. · ↓
26. · ↓
27. · 9.1
28. · ENTER
29. · 2ND
30. · STAT
31. · ↓
32. · ↓
33. · ↓

In 1-V mode each Y register is the frequency of the X above it and defaults to 1, so the double arrow-down leaves every return counted once. Set a Y to 3 and that return is weighted three times.

Register	Value	Meaning
n	6	Observations — check this first, always
X̄	3.1000	Arithmetic mean return
Sx	4.9751	Sample standard deviation, divisor n − 1
σx	4.5417	Population standard deviation, divisor n
ΣX	18.6	Sum, useful as a keying check

The rule is short: historical returns are a sample, so use Sx. Use σx only when the question states that you have the entire population — every month of a closed fund’s life, every employee in a firm, every bond in a defined index.

Why it matters downstream:

Using	Standard deviation	X̄ − 1.645σ	Answer choices land at
Sx (sample)	4.9751	−5.0841%	around −5.1%
σx (population)	4.5417	−4.3710%	around −4.4%

Seven-tenths of a percentage point apart on six observations, and the gap widens as n shrinks. Exam writers know this and set the distractors accordingly — the wrong deviation almost always has an answer choice waiting for it.

Three payments, all plausible, one correct

A £250,000 mortgage over 30 years at 6.5% nominal, paid monthly.

What was entered	PMT	Why it looks fine
P/Y 12, N 360, I/Y 6.5	−1,580.17	Correct
P/Y 12, N 30	−9,051.25	A number in the right shape for a payment
P/Y 1, N 360	−16,250.00	Also a number; also on the answer sheet

None of the three produces an error message. That is the whole problem with leftover settings — the calculator has no way to know you meant years when you typed 30 into a register measured in months.

Two defences. First, set P/Y to 1 permanently and enter the periodic rate at I/Y yourself, so a monthly problem is 360 N, 0.541667 I/Y and the setting never moves. Second, learn the sanity check: a monthly payment on a long mortgage sits between about 0.6% and 0.8% of the balance. £1,580 on £250,000 is 0.63%. £9,051 is 3.6%, which is a car loan, not a mortgage.

The full setup sequence, and the four different clears, are in how to use the BA II Plus.

The annuity-due shortcut that keeps you out of BGN

CFA questions use annuities due constantly — rent paid in advance, an insurance premium at the start of each year, a lease. The obvious route is 2ND BGN, 2ND SET, 2ND QUIT, and the obvious risk is forgetting to switch back.

There is no need to switch at all.

1. PV of an annuity due = PV of the ordinary annuity × (1 + i)

The same factor works for FV. Every payment arrives one period earlier, so the whole series is worth one period's interest more.

Fifteen annual payments of £12,000, discounted at 5%.

Route	Keys	Result
END then multiply	15 N, 5 I/Y, 12000 PMT, 0 FV, CPT PV, × 1.05	−130,783.69
BGN mode	2ND BGN, 2ND SET, 2ND QUIT, then the same five registers	−130,783.69

The ordinary annuity is −124,555.8965, and £124,555.8965 × 1.05 = £130,783.69 exactly. Three keystrokes versus five, no mode to leave on, and the multiplication doubles as a check: if the annuity due is not slightly larger than the ordinary annuity, something is wrong.

Load the cash flows once, read the worksheet twice

The CF worksheet feeds NPV and IRR from the same entries. Item sets that ask for both — and they usually do — cost one extra keystroke, not a second round of typing.

1. CF
2. · 2ND
3. · CLR WORK
4. · [flows]
5. · 2ND
6. · QUIT
7. · NPV
8. · 11
9. · ENTER
10. · ↓
11. · CPT
12. · …
13. · IRR
14. · CPT

Nothing is cleared between the two. You can also return to NPV, type a different rate and recompute, which is how you build a profile in seconds.

Two exam-specific uses of that. A profile — press NPV, new rate, ENTER, ↓, CPT, repeat — answers “at what discount rate does this project become unattractive” without a second thought. And the crossover rate, the discount rate where two projects tie, is the IRR of the differences between the two series: enter one series minus the other and press IRR. For the standard pair of a £100,000 and a £300,000 project it is 10.1783%, and it is the most efficient answer to any “at what rate are you indifferent” item. NPV vs IRR works the whole comparison through.

Read the answer the question actually asked for

The calculator gives you a number. The answer choices are frequently a different number derived from it, and the gap is where marks go missing.

Fractional N. How long to grow £250,000 into £1,000,000 at 6.5%? CPT N returns 22.0135. If the choices are years, the answer is 22.01. If the stem says “complete years”, it is 23 — you are not there at the end of year 22. If the stem says months, multiply by 12.

Periodic versus annual rates. An IRR computed on monthly flows is a monthly rate. Compounding is the conversion, not multiplication: 1.3888% a month is (1.013888)¹² − 1 = 18.00% a year, not 16.67%.

Sign. CPT PV on an inflow stream returns a negative number because of the cash flow convention. The answer choice will be positive. This is not a mistake to fix, it is a display convention to read past.

Never re-key an intermediate

Ten memories, STO 0 through STO 9, and RCL to bring one back. 2ND MEM lists all ten.

1. CPT
2. · PV
3. · STO
4. · 1
5. · …
6. · RCL
7. · 1
8. · ×
9. · 1.05
10. · =

2ND ANS recalls the last computed value, which covers the common case without spending a memory.

This is worth more than it sounds on multi-part items. A three-part vignette where part (a) is a bond price, part (b) applies a duration to it and part (c) asks for the new price is three uses of the same figure. Typed once, it is right or wrong once. Typed three times, it has three chances to be wrong, and a transcription error in part (a) silently poisons (b) and (c).

The same argument is why DEC = 9 matters. A price of 95.3994 rounded to 95.40 on paper and re-entered introduces an error at the fourth digit, which is exactly where the answer choices are separated on a yield question.

What to drill, and how fast

Computation	Keys	Target
Any single TVM unknown	Five registers, CPT	15 s
Nominal to effective rate	2ND ICONV	15 s
Mean and Sx from six numbers	2ND DATA, 2ND STAT	40 s
NPV and IRR from five flows	CF, NPV, IRR	45 s
Bond price from a yield	2ND BOND	40 s
Interest paid in a given year	2ND AMORT	25 s
Annuity due, via the ×(1+i) shortcut	TVM plus one multiply	20 s
Forward rate from two spot rates	y^x and ÷	25 s

That last row is worth a line of its own, because candidates reach for the formula sheet and lose thirty seconds: the one-year rate one year forward, given a two-year spot of 4.5% and a one-year spot of 3.8%, is 1.045² ÷ 1.038 − 1 = 5.204721%. It is four keystrokes and a division.

The morning of

1. Fresh batteries, and a second calculator. A dead battery is not a recoverable situation, and CFA Institute has historically permitted a spare approved model. Check the current policy yourself.
2. 2ND FORMAT 9 ENTER, 2ND P/Y 1 ENTER, confirm END. Thirty seconds, and it is the last time you touch a setting.
3. 2ND CLR TVM before every TVM item, 2ND CLR WORK inside every worksheet. Not 2ND RESET — that returns P/Y to 12 and decimals to two, mid-exam, with no undo.
4. Know which model is in your bag. Payback, net future value and modified IRR have registers on the Professional and workarounds on the standard model; the differences are here.

Speed on the keypad is the cheapest mark on the paper. It is entirely mechanical, it transfers to every level, and it is the one part of exam preparation where an hour of drilling has a guaranteed return. FRM candidates need a slightly different set — more volatility, more bond maths, the same keypad.