This is the beginning of Part 3 of our module on Measuring and Managing Counterparty Credit Risk, which examines in detail counterparty credit risk under OTC derivatives and shows you how to model this exposure for standard derivative contracts, in addition to reviewing common techniques for the mitigation of this risk. We remind you that this Part 3 contains two chapters, Chapter 3 which examines the marginal effect on an existing portfolio’s credit risk measures of a new transaction, and Chapter 4 which contains the final quiz. This Chapter 3 examines the impact of adding a new transaction to an existing portfolio’s credit profile.
It will demonstrate that the credit exposure from the new transaction, considered in isolation, may be quite different – and often substantially larger – than when it is aggregated with an existing portfolio. We begin with a portfolio that consists of one transaction only, our old $200MM IRS. To save the time of simulating and valuing the portfolio anew, we assume the current exposure is zero and the IRS remaining maturity is 5 years.
This enables us to use the same simulation results generated previously in WS Simulation in Excel file Simulation. xls. The transaction we add to the portfolio is the 3-yr FX forward we examined earlier, with the customer buying EUR50MM against USD.
We will compare the marginal impact of this new transaction on the portfolio peak PFE and expected exposure curves, versus on a standalone basis. We consider three cases: 1- no netting agreement exists, 2- enforceable netting is in place, and 3- enforceable netting is in place as well as a margining agreement with a $9MM threshold and monthly frequency of margin calls. The results appear in WS Comparison 1 of Excel file Portfolio.
xls. Cells B3:E4 list the details of the existing portfolio and the additional FX transaction. Cells B8:BK10 calculate the peak PFE for the existing portfolio assuming alternatively the absence of netting on this Row (Row 8), the availability of netting on this row (Row 9) and finally the combination of netting and margining on this row where the margin threshold (MT) is $9MM (Row 10).
In building this worksheet, we returned to WS Simulation in Excel file Simulation. xls, and reflected the portfolio under consideration here by changing the notional amount in that other worksheet for the excluded transactions to zero, specifically the 1-year and 3-year FX forwards. We next copied the future values from cells B3037:BK4539 into Cells B23:BK1526 of Excel file Portfolio.
xls, WS Netting & collateral. Additionally, we copied the peak PFE values and the expected exposure values and pasted them appropriately into this WS Comparison 1. We now assess the marginal impact of the new FX transaction on peak PFE, by calculating it first for the existing portfolio (Cells C8:BK10) and then recalculating it after including the new transaction (Cells C12:BK14) in the next table.
We then do the same for expected exposure in Rows 25: 27 and in Rows 29: 31. The marginal impact of the FX forward on peak PFE (Cells C16:BK18) appearing here and on expected exposure appearing further down over here (C33:BK35) is quite simply the difference between the old and the new measures. To calculate peak PFE and expected exposure for the new transaction on a standalone basis, we repeat the same steps as before, changing the notional amounts for the 1-year FX forward and for the IRS to zero.
The graphs to the right show the evolution of peak PFE and underneath expected exposure created by the transaction on both a standalone basis and when it is added to the existing portfolio. The results are very interesting: • When taken on a standalone basis with no credit mitigant, peak PFE is exactly the same as we had earlier, with a maximum peak PFE of $27. 9MM.
When the margining agreement is introduced, peak PFE quickly reaches the $9MM threshold and remains at this level until maturity. A similar pattern is visible in the graph for the expected exposure appearing now, where the graph rises steadily until maturity in the absence of any credit mitigant, but flattens in later periods when margining is introduced, since exposures in excess of $9MM are now replaced by the margining threshold. • Turning to the transaction’s marginal impact on the existing portfolio, we note the following: • Even in the absence of netting, peak PFE declines significantly.
This results from the imperfect correlation between the IRS and the FX forward. Specifically, when credit exposure under one instrument increases significantly, it increases under the other one only moderately if correlation is positive but low, and it even decreases if correlation is negative. Expected exposure, however, is the same in the standalone and portfolio case: here correlation has no impact on the average; remember that in the absence of netting, PFEs are simply added together.
• Netting further reduces peak PFE but only slightly, whereas its impact on expected exposure is much more significant – a result you should by now have anticipated: whenever exposures from two positions have opposite signs, expected aggregate exposure diminishes. • The scenario with both enforceable netting and the margining agreement is especially interesting. The impact on peak PFE is marginal in the first 5 months and zero after that, the reason behind that peak PFE has already exceeded significantly the margining threshold of 9MM, after 5 months, so any additional exposure from this stage onward will be fully collateralized – therefore, leaving credit exposure unchanged.
Expected exposure however is not zero here since on average the portfolio exposure lies below the margining threshold, so the addition of the new transaction that does not offset the existing ones on average increases credit exposure. The conclusion from this analysis is that the addition of the FX transaction to the existing portfolio has had a significant impact on the credit line required: this declines from around $28MM over 3 years on a standalone basis with no credit mitigation, to around $16. 5MM in the portfolio context with or without netting, and from $9MM over 3 years on a standalone basis to $2.
3MM over 6 months in the presence of a margining agreement with a $9MM margining threshold. Our second example assumes that the existing portfolio already includes two transactions, the IRS and the 3-year FX forward. Again we assume that current exposure is 0, and now add to this portfolio a new transaction, which is the 1-year FX forward.
We repeat the procedure from the preceding example on this WS, Comparison 2. We take you directly to the graphs to the right, which again reveal very interesting results: • On a standalone basis with no credit mitigant, peak PFE and expected exposure follow a pattern similar to the preceding example with maximum peak PFE at 9MM. The introduction of the margining agreement, appearing in orange, has virtually no effect on this, since PFE over the next 12 months lies at all times below the 9MM margining threshold.
The same effect is visible on the graph underneath, where the orange line once again matches more or less the original dashed line. • Adding the new transaction to the existing portfolio leads once again to interesting results: • In the absence of netting, peak PFE declines significantly, with a maximum around $2MM. As before, this is attributed to the imperfect correlation between the 1-year forward and the existing portfolio.
And as before, expected exposure is the same as for the standalone case. • In the presence of enforceable netting, the impact of the new transaction on credit exposure is actually beneficial – that is to say, booking the new transaction reduces both exposure measures as you can see from the dark blue line and this is true for both peak PFE and for expected exposure. In particular, the client is selling EUR under the 1-year forward, while buying EUR under the 3-year forward.
The two transactions offset each other virtually perfectly over the next 12 months. • The scenario with enforceable netting and the margining agreement produces similar results, shown in green, with a negative impact initially on peak PFE and also a negative impact all the time on expected exposure. Specifically in the case of peak PFE, the impact is beneficial for the first 5 months, when total portfolio exposure lies below the margining threshold, but then becomes 0 when the net effect is above the threshold and therefore collateralization is in place anyway.
Credit lines required to book the additional transaction have declined, from $9MM over 1 year in the standalone case with no credit mitigants, to around $2MM in the case without enforceable netting, and to zero 0 in the last case, with both netting and the margining agreement. The credit officer should very glad to approve this risk-reducing transaction. This completes this Chapter 3 and brings us to the final quiz.
