The uncertainties in the current market environ-ment both amplify existing risk factors and create new risks that have the potential to hurt whole loan valuations. Along with the obvious credit risk (driven by either individual or servicers’ behavior, HPA and other macro economic factors) other risks impose challenges to the investors’ decisions. Those risks include government activity (including modification); correlation with other markets (i.e., equities); and interest rate risk, inclusive of prepay risk. Hedging the interest rate risks, while being always relevant for agency product, has also become increasingly important in recent months for non-agency mortgage products. Moreover, with rates having been at their historical lows (and the likelihood of yields pushing higher after the QE2 era ends), the possibility that mortgages may sell-off must be given weight.
Although this article does not focus on any particular sector of the market, it is important to remark briefly on hedging needs in the non-agency sector. Before the crisis, most hedging of non-agency product centered on duration/convexity hedging, but when yields rose to the two-digit levels in 2009, market participants could disregard these risks. However, with yields now compressed to single digit levels for certain products (prime/jumbo), it is again becoming important to hedge against rate and prepayment risks. Although the correlation between prepayments and interest rates recently has eased for certain sectors, it still remains relevant for prime collateral.
Whole loan portfolio interest rate risk character-istics can be categorized into the following: duration (with curve exposure), convexity, volatility exposure and mortgage spread sensitivity. The hedging objective is the minimization of those risks. Mitigating these risks can be achieved through the use of swaps, as options can be used to eliminate residual exposure.
To demonstrate some of the hedging practices discussed here, we provide first the interest rate risk profile for a sample newly originated fixed rate and hybrid arm mortgage portfolio of $70M notional. Risk metrics were generated using the industry standard constant OAS methodology provided in the MIAC analytics suite and are shown in Figure 1.
With hedging the duration and curve risk as the starting point, we used model-generated durations and key-rate durations to arrive at the swap hedge amounts for the portfolio’s asset. Generally speaking, hedging both the long and short end of the curve is equivalent to hedging the overall duration and additionally applying a curve flattener. The different mix of swaps needed to hedge the curve exposure of different types of whole loans is based on the model. For example, given that the curve exposure of hybrid ARMs is skewed toward the short and intermediate end of the curve, the notional of 2-year swaps for a 3/1 ARM is understandably much higher than would be required for a 30-year fixed-rate mortgage. This is shown in Figure 2.
When the yield curve follows an upward slope, making whole loans attractive, it also makes hedging expensive. The challenge in hedging is to better understand the trade-off between performance and cost of carry. Hedging duration/curve risk involves selling 2-, 5-, and 10-year swaps; the steep curve makes this hedge expensive as the yield curve "rolls down" over time. For the sample portfolio above, just hedging the base case duration, using 2s, 5s and some 10s, costs approximately a point in return over 6-months. This is shown in Figure 3. In addition, this strategy has not addressed the negative convexity yet. At the same time, the substantial asset carry overwhelms the cost of hedging. Carry contributes over 15 points to the un-hedged return over a 6-month horizon. Comparing hedged and un-hedged returns over this horizon, the un-hedged asset outperforms hedged returns in scenarios up to +50bp. In the larger sell-off scenarios though, the benefits of hedging start to materialize. As can be seen in Figure 3, given a +100bp shift in rates, the hedged return beats the raw return by over 2 points.
Hedging to model duration (OAD) alone won’t be sufficient for mortgage portfolios with large negative convexity. Vega, the effect of fluctuations in implied volatility, and the gamma (convexity) exposure should be also taken into account. In general, hedgers employ options such as swaptions and their strategies (like straddles), allowing them to minimize those residual risks once the curve risk has been effectively hedged. In the strong rally or sell-off scenarios, as illustrated in Figure 4, hedging the same portfolio with options enhances convexity and improves hedging performance versus the swaps-alone strategy. However, the cost of hedging with options should also be considered and analyzed from a potential risk/return perspective. With option premiums trading historically 10%-15% above realized volatility, options may seem expensive. Hence, with rates being range bound, the use of options may not seem justified. In addition, other market factors beyond the scope of this article, such as volatility ratio (implied over realized), rate expectations and Fed dates, also should be considered when making decisions to execute options strategies.
Hedging strategies which use other popular hedge instruments such as amortizing swaps, index amortizing swaps (IAS), or Eurodollars should not be overlooked. The critical benefit of an index amortizing swap - where the fixed-rate payer has the option to amortize the notional principal of the swap if interest rates fall - is that it demonstrates positive convexity. Therefore, this swap can be used to hedge against the negative convexity exhibited by many whole loans and mortgage-backed securities characterized by high prepayments when interest rates decline. However, in today’s market environment, this might not be totally applicable to certain mortgage products because of their slow prepayments. In contrast, non-index amortizing swaps, are found to be an effective instrument for hedging non-agency products given their limited prepayment sensitivity to rates, when borrowers cannot find new loans, such as in the current environment.
Artem Lysenko
Senior Vice President, Capital Markets Group