Expected and Unexpected Inflation

Why Expectations Matter

Inflation is the rate at which the general price level rises over a stated period. Before that period ends, households, businesses, and analysts may form an expectation about the rate. Once prices are observed, the actual rate can be compared with the forecast. The difference matters because a nominal amount of money does not have fixed purchasing power: its value in goods and services depends on the price level that actually prevails.

For example, suppose someone plans using an expected inflation rate of 5%, but prices rise by 7%. The 7% rate is inflation; the difference between 7% and 5% is an inflation surprise, or forecast error. Keeping these quantities separate is essential when interpreting nominal payments, budgets, and purchasing power.

Definition: Expected inflation, written $\pi^e$, is the inflation rate forecast for a specified period using information available before that period is observed. Actual inflation, written $\pi^a$, is the rate measured for that same period after the relevant price data become available. The signed inflation forecast error is $\varepsilon=\pi^a-\pi^e$. A positive error means actual inflation exceeded the forecast; a negative error means it was below the forecast.

The comparison is meaningful only when the forecast and the observed rate refer to the same price measure and time interval. A forecast of annual CPI inflation should not be compared directly with a monthly inflation rate or with inflation measured by a different index. As in the earlier material on inflation rates and price indexes, a stated inflation rate summarizes a change in a price level over a defined period.

Forecast error is usually reported in percentage points. If expected inflation is 5% and actual inflation is 7%, then $\varepsilon=7\%-5\%=+2$ percentage points. It is not a 2% relative increase in the forecast. The relative increase from 5% to 7% would be 40%, but that is a different calculation and is generally not the measure intended when discussing an inflation surprise.

From Inflation to Purchasing Power

A nominal payment is an amount stated in dollars at the time it is paid. To interpret its purchasing power in the dollars of a base period, adjust it for the change in the price level. Assume the price index at the start of the period is $P_0$, and let $M$ be a nominal payment made at the end of the period. If the inflation rate over the period is $\pi$, the ending price index is $P_0(1+\pi)$. The payment’s purchasing power, measured in base-period dollars, is therefore $M P_0/[P_0(1+\pi)]$.

\text{Purchasing power in base-period dollars} =\frac{M}{1+\pi}

This formula assumes one inflation rate for the whole period, a consistent price index, and a rate greater than $-100\%$, so the price-level factor is positive. It does not say that every household experiences exactly this change in its personal cost of living: an index measures the price change for its defined basket and coverage. For a multi-year horizon, use the compounded price-level factor, not the sum of the annual rates.

Why an Inflation Surprise Changes Real Value

Suppose the same nominal payment $M$ is evaluated using a forecast and then using the price level that actually occurs. Under the one-period assumptions above, expected purchasing power uses $\pi^e$, while realized purchasing power uses $\pi^a$. Subtracting the two gives the change in purchasing power attributable to the forecast error.

\frac{M}{1+\pi^e}-\frac{M}{1+\pi^a} = \frac{M(\pi^a-\pi^e)} {(1+\pi^e)(1+\pi^a)}

The equality follows by taking a common denominator: the numerator becomes $M[(1+\pi^a)-(1+\pi^e)]$, which simplifies to $M(\pi^a-\pi^e)$. Because both price-level factors are positive, the sign of the difference is the sign of the forecast error. If actual inflation is higher than expected, the payment buys less than forecast; if actual inflation is lower than expected, it buys more than forecast. This conclusion concerns the payment’s purchasing power relative to the forecast, not whether the payment’s nominal dollar amount changed.

Key distinction: Expected inflation is a forecast, not a guarantee. Unexpected inflation is the difference between the actual rate and that forecast. A positive inflation rate can be fully expected, while an inflation surprise can be positive, zero, or negative.

Worked Examples

Worked Example: A One-Year Inflation Surprise

A hypothetical grant is expected to pay $\$2{,}100$ in one year. At the time of planning, the expected annual inflation rate is 5%. Inflation over the year is later measured at 7%. Find the forecast error and compare expected with realized purchasing power in today’s dollars. State the method and avoid confusing percentage points with a relative percent change.

Step 1: Find the forecast error. Use actual minus expected:

\varepsilon=7\%-5\%=+2\text{ percentage points}

Step 2: Calculate expected purchasing power. Deflate the nominal payment by the expected price-level factor:

\frac{\$2{,}100}{1.05}=\$2{,}000.00

Step 3: Calculate realized purchasing power. Use the actual inflation rate instead:

\frac{\$2{,}100}{1.07}=\$1{,}962.6168\ldots \approx \$1{,}962.62

The realized purchasing power is $\$2{,}000.00-\$1{,}962.62=\$37.38$ lower than expected. As a check, actual inflation exceeds expected inflation, so a fixed nominal amount must have lower purchasing power than forecast. The common mistake is to call the forecast error “2%” without specifying that it is 2 percentage points, or to apply 7% as a subtraction from the nominal payment rather than divide by the price-level factor $1.07$.

Worked Example: Compounding Forecast Errors Across Two Years

A hypothetical fund plans around annual inflation forecasts of 2% in year 1 and 3% in year 2. The rates later observed are 4% in year 1 and 3% in year 2. A nominal payment of $\$10{,}000$ is made at the end of year 2. Calculate the cumulative expected and actual inflation, the forecast error in each year, and the payment’s expected and realized purchasing power in today’s dollars. Use compounded price-level factors; do not add annual rates.

Step 1: Find the annual forecast errors.

\varepsilon_1=4\%-2\%=+2\text{ percentage points}, \qquad \varepsilon_2=3\%-3\%=0\text{ percentage points}

Step 2: Compound the expected annual rates. The expected two-year price-level factor is $(1.02)(1.03)=1.0506$, so cumulative expected inflation is $1.0506-1=0.0506$, or 5.06%.

Step 3: Compound the actual annual rates. The actual two-year price-level factor is $(1.04)(1.03)=1.0712$, so cumulative actual inflation is $1.0712-1=0.0712$, or 7.12%.

Step 4: Deflate the payment under each price-level path.

\text{Expected purchasing power} =\frac{\$10{,}000}{1.0506} =\$9{,}518.37

\text{Realized purchasing power} =\frac{\$10{,}000}{1.0712} =\$9{,}335.32

The realized purchasing power is $\$183.05$ lower than expected. The answer is consistent with the cumulative rates: the actual price-level factor, 1.0712, is greater than the expected factor, 1.0506. The method avoids treating the annual rates as additive: adding the expected rates would give 5%, not the correct compounded rate of 5.06%.

Worked Example: Comparing Price-Index Forecasts

A hypothetical price index is 240 at the start of a year. A forecast puts the index at 252 at year-end, but the observed index is 264. A nominal payment of $\$2{,}520$ is made at year-end. Find forecast inflation, actual inflation, the forecast error, and the payment’s expected and realized purchasing power in base-period dollars. Use the index ratios rather than assuming the index begins at 100.

Step 1: Calculate each inflation rate from the initial index.

\pi^e=\frac{252-240}{240}=0.05=5\%, \qquad \pi^a=\frac{264-240}{240}=0.10=10\%

Thus, $\varepsilon=10\%-5\%=+5$ percentage points.

Step 2: Calculate the payment’s expected purchasing power. Convert the forecast end-year amount into base-period dollars:

\$2{,}520\left(\frac{240}{252}\right)=\$2{,}400.00

Step 3: Calculate realized purchasing power. Use the observed end-year index:

\$2{,}520\left(\frac{240}{264}\right) =\$2{,}290.9090\ldots \approx \$2{,}290.91

The realized amount is $\$109.09$ below the expected amount. The same result follows from dividing by 1.10 rather than 1.05. The check is that the actual index rose more than forecast, so each end-year dollar represents less base-period purchasing power. A common mistake is to calculate inflation as the change in the index, 12 or 24 index points, without dividing by the starting index of 240.

Interpreting Forecast Errors Carefully

An inflation forecast can be wrong even when it is reasonable given the information available when it was made. The error is identified after comparing the forecast with the observed outcome; it does not, by itself, establish why the forecast differed from reality. For example, a surprise may reflect new information about demand, costs, or other influences on prices. The calculation of the error does not identify the cause.

A negative forecast error does not necessarily mean prices fell. If expected inflation is 6% and actual inflation is 3%, the error is $-3$ percentage points, but the price level still rose. Prices fall only when the actual inflation rate is negative. If expected inflation is 3% and actual inflation is $-1$%, then the error is $-4$ percentage points and the actual price level fell over the period.

For multi-period comparisons, the relevant cumulative price-level change is the product of the annual factors. Even if annual forecast errors partly offset, the accumulated expected and actual price levels should be compared using their compounded factors. The effect on a nominal payment depends on those cumulative levels, not simply on the sum of the annual forecast errors.

Common Mistakes

Mixing up actual inflation and the forecast error. If expected inflation is 4% and actual inflation is 6%, actual inflation is 6%; the error is +2 percentage points.
Using the wrong sign. Define the error consistently as actual minus expected. Reversing the order reverses its interpretation.
Subtracting the inflation rate from a nominal amount. To express a future nominal payment in base-period dollars, divide by the price-level factor, such as $1+\pi$, or use the corresponding price-index ratio.
Adding rates across years. Compound the annual price-level factors. For rates $r_1$ and $r_2$, the two-period factor is $(1+r_1)(1+r_2)$.
Assuming a negative surprise means deflation. A negative error only says actual inflation was below expected inflation. Actual inflation must itself be negative for the price level to fall.

Key takeaway: Compare actual and expected inflation over the same period and measure, using $\varepsilon=\pi^a-\pi^e$. To assess a nominal payment’s purchasing power, adjust it by the price-level factor implied by the relevant forecast or observed inflation. A positive surprise lowers realized purchasing power relative to what was expected; a negative surprise raises it.

Check Your Understanding

For each problem, state the forecast error where requested and use the appropriate price-level factor to interpret purchasing power.

A hypothetical payment of $\$1{,}575$ is expected one year from now. Expected inflation is 4%, and actual inflation is 6%. Find the forecast error and the expected and realized purchasing power in today’s dollars. (Answer: $+2$ percentage points; expected $\$1{,}514.42$, realized $\$1{,}485.85$.)
A price index is currently 180. The forecast index one year later is 189, while the actual index is 198. A hypothetical nominal payment is $\$1{,}980$. Find forecast inflation, actual inflation, and realized purchasing power in current-period dollars. (Answer: expected inflation $5\%$, actual inflation $10\%$, realized purchasing power $\$1{,}800.00$.)
Expected inflation is 1% in year 1 and 2% in year 2; actual inflation is 2% in year 1 and 2% in year 2. For a hypothetical $\$5{,}000$ payment at the end of year 2, find the two cumulative inflation rates and its expected and realized purchasing power in today’s dollars. (Answer: cumulative expected inflation $3.02\%$, cumulative actual inflation $4.04\%$; expected purchasing power $\$4{,}853.43$, realized purchasing power $\$4{,}805.84$.)
Expected inflation is 3%, but actual inflation is $-1$%. A hypothetical year-end payment is $\$990$. Find the forecast error and the payment’s realized purchasing power in today’s dollars. (Answer: $-4$ percentage points; realized purchasing power $\$1{,}000.00$, with actual deflation.)

What You'll Learn