Sources and Types of Financial Risks

1 Introduction

Section C of the syllabus deals with financial risk: the risk that an entity's cash flows, its reported profit or its value are changed by movements in financial variables it does not control. This chapter identifies those risks and shows how one of them can be measured. The two chapters that follow deal with managing them — chapter 9 with currency risk, chapter 10 with interest rate risk.

Three questions run through all three chapters, in order:

1. Where does the exposure come from, and what kind of risk is it? (Sections 2 and 3 below.)

2. How big is it? (Section 5 below, and the exchange rate forecasting in chapter 9.)

3. What, if anything, should be done about it? (Section 4 below for political and economic risk; chapters 9 and 10 for currency and interest rate risk.)

2 The sources of financial risk

Financial risk is not something an entity chooses to take on separately from its business. It is created by ordinary commercial and financing decisions, and the exposure can usually be traced back to the decision that produced it.

2.1 Which decision creates which exposure

Two consequences follow, and both are examinable.

• An entity can be exposed without ever dealing in a foreign currency. If its competitors import, a movement in the exchange rate changes their prices and therefore its own market share. That is economic risk and no invoice of the entity's own is involved.

• The exposure is created before the risk is visible. A quotation given in a foreign currency creates an exposure at the moment it is accepted, not at the moment the invoice is raised; a decision to fund a project with floating rate debt creates one for as long as the debt is outstanding.

2.2 Cash flow risk and value risk

It is worth separating two things the word "risk" is doing, because the responses to them differ.

2.3 Deciding to bear the risk

Hedging is not free. A forward contract commits the entity irrevocably, an option costs a premium, and every hedge costs management time and creates a position that has itself to be monitored. Where the exposure is small relative to the entity, or one-off, or where the entity has a natural offset elsewhere in the group, the correct recommendation may be to accept the exposure and say so. What is never acceptable is to leave a material exposure unmeasured.

3 The types of financial risk

The syllabus names four: economic risk, political risk, currency risk and interest rate risk.

3.1 Interest rate risk

An entity may borrow, or deposit, at either a floating rate or a fixed rate.

• At a floating rate the interest is reset periodically by reference to a published benchmark rate, so the cash flows are uncertain for as long as the borrowing lasts. A rise in rates increases the cost and reduces cover; a fall reduces it.

• At a fixed rate the cash flows are certain. That is not the same as being free of risk: the entity is locked in, so if rates fall it carries a cost its competitors no longer bear, and the market value of the debt rises against it.

There is a third case, and it is the one examiners like, because the exposure exists before any borrowing does.

3.2 Currency risk

Any transaction, asset or liability denominated in a currency other than the one the entity reports in exposes it to movements in the exchange rate. Currency risk divides into three, and the division matters because the three call for different responses.

All three, and the instruments used against them, are the subject of chapter 9.

3.3 Economic risk

The syllabus lists economic risk separately from currency risk, and the wider reading is the right one. Economic risk is the exposure of an entity's future cash flows to the economic environment as a whole: the rate of inflation, the level of interest rates, the rate of growth in the markets it sells into, commodity and energy prices, and — through all of these — the level and pattern of demand.

Two features distinguish it from the other three risks:

• it affects future, uncommitted business rather than a transaction already entered into, so there is no single amount and no single date to hedge;

• it is largely undiversifiable by financial instruments, which is why the responses to it in section 4 and in chapter 9 are operational — where the entity makes things, where it buys, which currencies it invoices in — rather than contractual.

Chapter 1 works the quantitative side of this: an example that re-forecasts a statement of profit or loss for predicted changes in sales volumes, costs, interest rates and the tax rate, and measures the effect on earnings.

3.4 Political risk

Political risk is the risk that action by a government, or instability in the country concerned, changes the position and value of the entity.

It is conventionally split by how widely the action falls.

4 Managing political and economic risk

There is no market in which political risk can be bought and sold, so the responses to it are commercial rather than contractual. They fall into four families, and the first of them is by far the most effective because it happens before the money is committed.

4.1 Negotiate the environment before investing

• Agree the terms of the investment with the host government in advance — what may be remitted, what tax will be charged, what local participation is required, and what happens if the assets are taken.

• Obtain insurance against expropriation and against inconvertibility, either privately or through an export credit agency in the entity's home country.

• Secure local government support: grants, tax holidays, infrastructure, and a stake in the project's success.

4.2 Operating, marketing and financial responses

4.3 Staging and exit

Two further responses belong to the investment decision rather than to the treasury function: commit the investment in stages, so that each further tranche is a fresh decision taken with better information, and keep the exit cheap — lease rather than build, and avoid sinking capital into assets that cannot be moved or sold, where the political assessment is uncertain.

4.4 Managing economic risk

Economic risk is managed in the same way, and for the same reason: no contract can hedge a competitive position. The responses are to spread the exposure across more than one currency zone and more than one market, so that the movements do not all fall the same way at the same time; and to match the currency of costs to the currency of revenues by producing, buying and borrowing where the entity sells. Chapter 9, section 11, works this through for currency.

5 Quantifying financial risk: value at risk

Identifying an exposure is not the same as measuring it. Value at risk is the standard technique for putting a single number on the risk in a position whose value moves from day to day — a shareholding, a portfolio, a trading book, or a foreign currency balance.

5.1 The normal distribution

The technique assumes that the value of the investment varies around a mean value following a normal distribution. That assumption is what allows a probability to be attached to any outcome.

If the possible results cluster closely around the mean, the standard deviation is small; if they are widely spread, it is large.

So if the mean daily value of a shareholding is $30 and the standard deviation of its value is $1, the holding is rarely valued far from $30. If the standard deviation were $10 the value would be very volatile — often more than $40 and often less than $20, on the same mean.

5.2 The Z score and the normal distribution table

All normal curves have the same shape, so one table serves all of them. The table is entered with Z, the distance of a value from the mean expressed as a number of standard deviations:

Z = (x − μ) ÷ σ

where x is the value, μ the mean and σ the standard deviation. The body of the table gives the proportion of all outcomes lying between the mean and Z standard deviations from it, on one side only.

For example, if the mean height of a population is 178 cm with a standard deviation of 4 cm, the proportion of the population between 178 cm and 181 cm tall is found from:

Z = (181 − 178) ÷ 4 = 0.75

Read the table down the left-hand column to 0.7 and across to 0.05: the figure is 0.2734. So 27.34% of the population is between 178 cm and 181 cm tall — and, because the curve is symmetrical, the same proportion is between 175 cm and 178 cm.

5.3 Reading the table backwards

The table is used in the other direction just as often: given a probability, find the value. Suppose the question is the height range containing the 20% of people who are just taller than the mean. The area required is 0.2, so look for 0.2 in the body of the table. It lies between Z = 0.52 (0.1985) and Z = 0.53 (0.2019), and 0.2 is almost midway, so Z is taken as 0.525. Then:

0.525 = (x − 178) ÷ 4, so x − 178 = 4 × 0.525 = 2.1

So the 20% of people just taller than the mean are between 178 cm and 180.1 cm tall.

5.4 Value at risk over one period

The commonest criterion is the amount that could be lost over a period such that there is only a 5% chance of losing more — that is, value at risk at the 95% confidence level. On the curve, the 5% is the left-hand tail:

The table gives areas measured from the mean, so the 5% tail is found by looking up the 45% that lies between the mean and the cut-off. An area of 0.45 corresponds to Z = 1.645 (midway between 1.64 and 1.65).

Suppose a shareholding has a mean value of $80,000 and the daily standard deviation of its value is $5,000. Below what value would only 5% of outcomes lie?

1.645 = (80,000 − x) ÷ 5,000, so x = 80,000 − (5,000 × 1.645) = $71,775

There is a 5% chance that after one day the holding will be worth less than $71,775, and a 95% chance that it will be worth more. The value at risk at the 95% confidence level is the maximum loss with 95% confidence:

Or, in one step, because the mean cancels:

VaR = Z × σ = 1.645 × $5,000 = $8,225

5.5 Changing the confidence level

At the 99% confidence level the curve is split 1% / 49% / 50%, so the area to look up is 0.49. The nearest table figure is 0.4901, at Z = 2.33.

VaR = 2.33 × $5,000 = $11,650

So on the same holding there is a 1% chance of losing more than $11,650 in a day. A higher confidence level always gives a larger value at risk on the same position: demanding to be right more often means allowing for a worse day.

5.6 Value at risk over several periods

The figures above are for one day, because the standard deviation used was a daily one. For a longer period the standard deviation has to be scaled, and the rule — which has to be learned — is that it scales with the square root of the number of periods, not with the number of periods:

σ_period = σ_day × √n

where n is the number of days in the period. So on a daily standard deviation of $5,000, the standard deviation over 10 days is:

$5,000 × √10 = $5,000 × 3.1623 = $15,811

and the value at risk over 10 days at the 95% confidence level is:

1.645 × $15,811 = $26,009

The ten-day figure must be larger than the one-day figure, because there could be a run of bad days. It is not ten times larger, because the good days and the bad days partly cancel: ten days of risk costs √10 ≈ 3.16 times the one-day figure, not 10 times.

5.7 What value at risk does not tell you

Value at risk is a single number and it buys its simplicity at a price. Every one of the following is a fair criticism in an exam answer:

• It says nothing about how bad the bad case is. A 95% figure of $8,225 says that 5 days in 100 will be worse; it does not say whether they will be worse by $1 or by $500,000. The measure that answers that question is the average loss beyond the cut-off, sometimes called expected shortfall.

• It assumes a normal distribution. Financial returns have fatter tails than the normal curve: extreme days happen more often than the model says, which means value at risk understates the risk exactly where it matters most.

• It assumes the standard deviation is stable and known. In practice σ is estimated from past data, and volatility itself rises in a crisis — so the measure is at its least reliable at the moment it is most needed.

• It is a statement about normal conditions, not about a crisis. It should be used alongside stress testing and scenario analysis, which ask what a specific extreme event would do, rather than what a probability distribution says.

• Adding up is not straightforward. The value at risk of two positions together is not the sum of their separate figures unless they are perfectly correlated; it is usually less, and how much less depends on a correlation that also moves.

6 Normal distribution table

The body of the table gives the proportion of outcomes lying between the mean and Z standard deviations from the mean, on one side of the curve.

7 Test your knowledge

Two quick checks before you move on: work through the flashcards to fix this chapter’s terms and definitions, then sit the objective questions for exam-style practice. Both mark themselves and explain the answers as you go.