Why Diagnose Hedge Outcomes?
A hedge can reduce an exposure without eliminating it, and a favorable or unfavorable result over one interval does not by itself establish whether the hedge was well designed. The useful first step is to reconcile the change in the exposure with the change in the hedge over the same interval, using a consistent valuation and sign convention. The residual then provides a starting point for diagnosis: it may reflect an intentional or unintentional mismatch, basis movement, timing differences, or an error in the data or model.
This tutorial concerns realized hedge outcomes, not the construction of a dynamic hedge or a full stress-testing framework. The earlier material on basis risk provides background on proxy relationships; here the emphasis is on interpreting the outcome and testing explanations against observed positions, dates, and market data.
Use changes on a consistent basis. For example, include relevant cash flows, settlement amounts, and accrued carry consistently in both valuations, or explicitly state that the calculation uses mark-to-market value changes only. Convert amounts to the same currency and compare the same start and end times. If the exposure is a liability, define its value change using the signed value of that liability; do not silently switch between a liability increase and a loss to the insurer.
The offset ratio has a directional interpretation. If exposure and hedge changes have opposite signs, \(O\) is positive: the hedge offsets some or all of the exposure movement. If they have the same sign, \(O\) is negative: the hedge reinforces that movement. A ratio between 0 and 1 indicates a partial offset in the observed interval; \(O=1\) indicates an exact offset; and \(O>1\) indicates that the hedge more than offsets the exposure change. These descriptions are specific to the measured interval, not proof that the hedge will behave similarly in another interval.
Deriving the Residual and Offset Interpretation
The combined change follows directly from adding the changes in the two positions. The offset ratio expresses the hedge change as a fraction of the exposure change, with a minus sign so that an exactly offsetting hedge has a positive ratio of one.
Rearranging the second equation gives \(\Delta V_H=-O\Delta V_E\). Substituting this into the first equation gives the residual identity:
Thus, if \(O=1\), the net change is zero; if \(0<O<1\), the residual has the same sign as the exposure change but is smaller in magnitude; and if \(O>1\), the residual has the opposite sign because the hedge has over-offset the exposure. If \(O<0\), the hedge moved in the same direction as the exposure, so the magnitude of the residual exceeds the magnitude of the exposure change. When \(\Delta V_E=0\), the ratio is undefined. In that case, report the hedge change and net change directly rather than forcing an offset ratio.
Worked Examples
Worked Example: Recognizing an Over-Offset
A hypothetical insurer’s asset exposure increases in value by $50,000 over a one-month interval. A hedge position loses $65,000 over exactly the same interval. Calculate the combined change and the offset ratio, then interpret the result.
Method. Keep the reported changes signed: \(\Delta V_E=+\$50{,}000\) and \(\Delta V_H=-\$65{,}000\). The changes have opposite signs, so the hedge moved opposite to the exposure.
Combined change.
Offset ratio.
The positive ratio confirms an offsetting direction, and its value above one shows that the hedge more than offset the exposure gain. The combined position lost $15,000. Common mistake avoided: reporting the ratio as \(-1.30\) because the hedge change is negative. The definition includes a leading minus sign; opposite-signed changes produce a positive offset ratio.
Worked Example: Separating a Proxy Basis Residual
A hypothetical insurer has an exposure whose value increases by $80,000 when its relevant market index rises by 4.0%. It hedges with a proxy instrument. For this position, the calibrated hedge sensitivity is a value decrease of $20,000 for each 1.0% rise in the proxy index. Over the same interval, the proxy index rises by 3.0%. Assume the exposure response and hedge sensitivity are linear over these changes, and exclude carry, cash flows, and other market drivers. Calculate the realized hedge change, the residual, and the part of the residual attributable to the difference between the two index moves under these assumptions.
Step 1: Calculate the realized hedge change. The proxy index rises by 3.0 percentage points, so the hedge change is three times its stated change per percentage point:
Step 2: Calculate the realized residual and ratio.
The hedge offset 75% of the exposure increase in this interval. To isolate the stated proxy mismatch, calculate the hedge change that would have occurred if the proxy index had also risen by 4.0%:
At that matched move, the residual would be zero under the assumptions. The proxy index instead rose by 1.0 percentage point less, leaving a $20,000 smaller hedge loss. That difference explains the $20,000 residual in this simplified scenario. Common mistake avoided: calling every incomplete offset a model error. Here the inputs and arithmetic are stipulated as correct; the observed difference between the exposure index and proxy index is the basis explanation.
Worked Example: Correcting a Position-Input Error
A hypothetical insurer’s exposure loses $100,000 over an aligned interval. A hedge instrument gains $1,000 per unit for the observed market move. The report shows a hedge gain of $70,000 because the model used 70 units. A position reconciliation verifies that the insurer actually held 100 units throughout the interval. Calculate the reported residual, then recalculate it using the verified position and diagnose the discrepancy.
Step 1: Reproduce the reported hedge change. The model’s 70-unit input gives:
The reported net change is \(-\$100{,}000+\$70{,}000=-\$30{,}000\). The reported offset ratio is \(-(+\$70{,}000)/(-\$100{,}000)=0.70\).
Step 2: Calculate the corrected hedge change from the verified units.
Using the verified position, the corrected combined change is \(-\$100{,}000+\$100{,}000=\$0\), and the corrected offset ratio is \(-(+\$100{,}000)/(-\$100{,}000)=1.00\). The observed $30,000 residual in the report came from the incorrect position input, not from evidence that the actual hedge under-offset the exposure. Common mistake avoided: treating the reported $70,000 as the hedge’s verified realized change. It is the result produced by the model’s 70-unit input; the corrected change from the reconciled 100 units is $100,000.
Worked Example: Finding a Valuation-Window Mismatch
A hypothetical insurer records an exposure gain of $40,000 from Monday’s opening valuation to Friday’s close. A preliminary report compares it with a hedge loss of $40,000 measured only from Monday’s opening valuation to Thursday’s close. The hedge then loses a further $15,000 between Thursday’s close and Friday’s close. Calculate the preliminary residual and the correctly aligned residual. Assume all figures are on the same currency and valuation basis.
Preliminary, misaligned comparison. Comparing the full exposure interval with only part of the hedge interval gives \(+\$40{,}000-\$40{,}000=\$0\). That apparent exact offset compares different end dates.
Aligned hedge change. Add the hedge’s final-day change to its Monday-to-Thursday change:
The Friday-close residual is therefore \(+\$40{,}000-\$55{,}000=-\$15{,}000\), and the aligned offset ratio is \(-(-\$55{,}000)/(+\$40{,}000)=1.375\). The report’s zero residual was an artifact of mismatched valuation windows. The aligned result is an over-offset for this interval; further investigation would be needed to explain the hedge’s additional loss. Common mistake avoided: interpreting a zero residual as evidence of a successful hedge before confirming that both changes cover the same dates.
A Practical Diagnostic Sequence
A residual is a measurement, not an explanation. Use a reconciliation sequence before concluding that the hedge design failed. Basis, timing, and model or implementation errors can coexist, so a diagnosis should be supported by evidence and should not force the outcome into only one category.
Confirm that exposure and hedge changes use the same start and end times, currency, valuation basis, and treatment of cash flows, settlements, and carry.
Verify actual quantities, signs, trade dates, maturities, and any changes in holdings. Compare those records with the position inputs used in the reported calculation.
Calculate \(\Delta V_{\mathrm{net}}\) and, when defined, \(O\). If the reported result cannot be reproduced, investigate arithmetic, sign, mapping, and data-processing errors before interpreting hedge effectiveness.
Identify which market factors moved the exposure and hedge. A difference between a proxy factor and the exposure’s relevant factor may create basis residuals even when both valuations and positions are correct.
Apply the verified positions and observed market data to the stated valuation method. A difference between that reproduced result and the reported outcome points toward a data, implementation, or model issue; determine which by examining the inputs and method.
What the Residual Does—and Does Not—Show
A correctly measured residual identifies the realized net change for the chosen positions and interval. It does not, by itself, prove whether the hedge was well designed. For instance, a positive offset ratio below one can reflect an intentionally partial hedge, a proxy mismatch, a market move outside the hedge’s intended scope, or an error in the calculation. Diagnosis depends on the hedge objective and the supporting records.
The offset ratio is also sensitive to its denominator. If the exposure barely changes, a small hedge movement can produce a very large ratio; if the exposure change is zero, the ratio cannot be calculated. In those circumstances, show the dollar changes and describe the market context rather than treating the ratio as a stable performance measure. Even when the exposure change is substantial, one realized interval is evidence about that interval, not a guarantee of future performance.
A common exam trap is to interpret the sign of \(\Delta V_H\) alone. A negative hedge change does not automatically mean the hedge failed: it offsets a positive exposure change, but reinforces a negative exposure change. Always retain the signs of both changes, calculate their sum, and then interpret their relationship. This disciplined reconciliation is aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus.
Check Your Understanding
For each problem, keep the value changes signed and check that the exposure and hedge intervals match before interpreting the result.
- A hypothetical portfolio exposure loses $80,000 while its hedge gains $60,000 over the same interval. Calculate the net change and offset ratio. (Answer: net change \(-\$20{,}000\); offset ratio \(0.75\); partial offset.)
- An exposure gains $45,000 and its hedge loses $54,000 over the same interval. Calculate the net change and offset ratio, and state whether the hedge under-offset or over-offset the exposure. (Answer: net change \(-\$9{,}000\); offset ratio \(1.20\); over-offset.)
- An exposure gains $64,000 when its relevant index rises by 4.0%. A proxy hedge loses $12,000 per 1.0% rise in its index, which rises by 4.5%. Assuming linear sensitivity and no other value changes, calculate the hedge change, net change, and offset ratio. (Answer: hedge change \(-\$54{,}000\); net change \(+\$10{,}000\); offset ratio \(0.8438\).)
- An exposure loses $72,000. A hedge gains $2,000 per unit for the market move. The report used 18 units, but position records verify 24 units. Calculate the reported and corrected net changes, and identify the cause of the difference. (Answer: reported hedge gain \(+\$36{,}000\) and reported net change \(-\$36{,}000\); corrected hedge gain \(+\$48{,}000\) and corrected net change \(-\$24{,}000\); the difference is the erroneous position input.)
- An exposure gains $30,000 from Monday’s opening valuation to Friday’s close. The hedge loses $20,000 from Monday’s opening to Thursday’s close and a further $5,000 from Thursday’s close to Friday’s close. Calculate the aligned net change and offset ratio, and explain the error in using only the first hedge interval. (Answer: aligned hedge change \(-\$25{,}000\); net change \(+\$5{,}000\); offset ratio \(0.8333\); using only the first interval would produce a misaligned net change of \(+\$10{,}000\).)