The nature of premium illusion in finance – Part I
Over the past three months I’ve devoted blog space to describing the nature and benefits of robust frameworks in the financial markets context. Changing tack from frameworks to a related issue, I’m going to write a series of blogs that examine an issue that those faced with market risk decision-making may not realise they encounter: the problem of premium illusion.
In doing so, I’m going to attempt to show that premium illusion is quite real, and why it can be a serial problem for those whose hedging frameworks exclude the use of option-based products.
To make these pieces more digestible I’m going to look at three simple cases that to me exemplify the problem, starting with the extended period of seemingly forever falling interest rates in the Australian interest rate markets. I want to stress up front that I chose, ex-ante, the Australian short-end rate market before I conducted any tests for actual evidence of premium illusion.
Falling rate persistency
As of this blog’s publication it is 4,131 days since the Reserve Bank of Australia last moved to raise the RBA Official Cash rate. That is, official rates have moved ratchet-style in only one direction for more than eleven years.
Mirroring these official moves, key interest-rate (IR) markets have moved in generalised lock-step with the prevailing cash rate, albeit with the kind of advance-and-lag process experienced dealers will be familiar with. This includes the recent heightened expectation of looming cash rate rises.
For all intents and purposes rates have been a one-way ‘bet’ for at least nine of the past ten years, and up until quite recently it has only rarely paid to hold outright paid-fixed IR-derivative positions.
For those tasked with interest-rate hedging, traditional IR- Swap programs that seek opportunities to fix forward rates have proved a burden throughout the seemingly never-ending decline in yields.
How can we be certain?
Here I will leave aside comparisons of floating-reset realised rates versus comparable spot-starting IR-Swaps for like tenors. This is because the nature of ultra-low short-term yield curves in the relevant period makes this a wholly unfair comparison.
Instead, I will look at realised rates of 1-Year into 1-Year (1y1y) forward-starting swaps versus comparable spot swap rates at each forward start date going back ten years.
Terminal Payoffs
Consider the case of a treasury dealer who has been required to periodically lock-in forward IR hedges via paid AUD 1y1y IR-Swaps.
Acknowledging that this is analysis by hindsight, it is still instructive to ask: how did this simple 1y1y forward hedging strategy perform against doing nothing (then swapping fixed via vanilla spot-swaps)?
The answer is, not well at all; the fall in rates was simply too persistent for forward paying to be of benefit.
Of 2,363 observed AUD 1y1y paid swaps, the realised outcomes since 1st February 2012 were:
In other words: 87.8% of the time the 1y1y forward rate paid exceeded the eventual 1y spot swap.
To summarise: IR-Swap hedging programs designed to periodically fix borrowing rates proved overwhelmingly costly versus simply doing nothing in the past ten years.
There was a persistent realised premium associated with forward hedging when we apply the following simple formula to the historic data:
What of comparable 1y1y payers swaptions?
For those unfamiliar with swaptions, an IR-Swaption is simply an option that gives the holder (i.e., the buyer) the right but not the obligation to enter into either a paid-fixed, or received-fixed IR-Swap (or a cash settlement amount equivalent to the intrinsic value of the swaption at the moment of expiry).
For all intents and purposes, swaptions markets operate like any other market for options, and OTC swaptions in major derivatives trade in $ trillions each year. They are quite liquid.
The standard pay-off for a swaption is quite similar to other such instruments.
For the purposes of display, I’ve created a hypothetical 1y1y PAYERS Swaption whose attributes I’ve made up from the rounded-up averages of available market data (May 2013, to the present):
Expiry: 1-Year
Underlying Swap: 1-Year Swap
Strike: 1.60%, i.e., at-the-money (ATM)
Premium: 0.45%
The counterintuitive
Having been involved in various options markets for many years I’m convinced that standard option “pay-off” diagrams that adorn the texts have served to mislead those who might otherwise use them as hedging instruments. In part, this has led to a chronic misunderstanding of the reason one should consider using option-based products when circumstances might make such consideration reasonable
Why?
Somewhat counterintuitively, the main benefit of option products when deployed as hedging instruments (not as expressions of trading views) is actually when the market moves against the prevailing strike, i.e., when the option FAILS to pay off.
Payoff diagrams serve to encourage the view that option buyers benefit only when the option moves in the direction favourable to the strike, but here I urge readers to think in terms of relative realised costs compared to fixed-rate alternatives.
Taking the example of our 1.60% PAYERS Swaptions, consider that:
1. If rates rise, the swaption protects the borrower at 1.60% for 0.45%, for an effective all-in swap of 2.05% fixed – the swaption can at no point ‘beat’ a paid-fixed swap at rates above the swaption strike price.
2. If rates fall, the borrower holds the option to let the swaption lapse and PAY swap rates at favourable levels (and if rates fall very quickly the borrower might PAY on swap at a favourable moment and sell back the swaption for whatever extrinsic value as can be recouped).
Point 2 almost perfectly describes the latitude the swaption actually provides which theoreticians call: intertemporal choice, which is a rather fancy concept that means the holder has time to pick a more favourable rate if one emerges. It describes the value of time in options products, affording holders freedom to choose favourably among given market states.
This is a long-underappreciated feature of option-based products, and associated strategies.
The 1y1y PAYERS Swaption versus 1y1y PAID Swap?
Over the slightly shorter 2,037 trading days (7 years, 9 months) for which we have swaption pay-off data what should be clear is that the swaptions performed similarly poorly to outright swaps. The rate markets rallied; rates went down throughout, both PAID Swaps and PAYERS Swaptions lost money outright.
But what about relative outcomes?
It turns out that PAYERS Swaptions outperformed PAID swaps with a not inconsiderable frequency:
To clarify
Let me pause here to recap and clarify what the data is telling us:
1. In the prevailing interest rate environment since 2012 forward IR hedges of all types proved costly,
2. Hedges proved costly almost 90% of the time (87.8% in the 1y1y)
3. The average hedging cost was 45.6 BP worse than doing nothing.
4. The alternative of PAYER Swaption hedges would have reduced this cost around 40% of the time – and likely more (given intertemporal choice), had they been used in place of PAID IR-Swaps.
It’s important to note that I’m comparing relative terminal outcomes between competing products, not expressing a view of outright gains or losses, nor am I seeking to push a product bias. As we will see in my next blog, there have been times when option-based hedges have been serially wasteful.
For those who maintained a hedge program, there are questions that could be asked:
§ Could deeper product appreciation of the potential benefits of optionality be deployed to your advantage?
§ Have you developed an approach to assessing whether fixed or options-based products are better tailored to your view or the prevailing environment?
How can we tell there is premium illusion?
In the market history I selected for this case study (1y1y AUD forwards, selected ex-ante) it’s clear that swaptions created opportunities for better hedge rates for vanilla swaps with good strong frequency, but I should concede I had an advantage in selecting the data, having traded my first option in 1989. The persistent fall in rates is a classic case for use of option-based hedges in preference to fixed hedges, and of course I knew this.
Yet, despite the 38.1% edge we find in the historic data, and the fact that option-based hedging instruments were clearly better tailored to a prevailing rate-environment, it’s not clear that the advantages hidden away behind pay-off diagrams are well known in the wider financial markets community.
So, I ask this question: if option-based hedges can be shown to be valid market instruments, why are they so infrequently used in hedging interest rate risk?
We know from long experience, from Depository Trust & Clearing Corporation (DTCC) data, and from elsewhere, that the use of options as hedges is dwarfed by fixed hedges across almost all markets.
The BIS data shows the poor take-up of option products emphatically:
Proportional market use by product-types:
Source: BIS
To our knowledge the use of interest-rate-options in global financial markets has at no time been above 10% of total USD derivative product usage, and in some jurisdictions the usage has been even lower. Worse, in our recent experience we have seen nothing to suggest any higher degree of interest in IR-Option hedges across markets, despite the volatility associated with COVID-19 and heightened uncertainty.
Which suggests there is something fundamental which deters people from deploying these products.
There are many possible factors at play here, some of which I will explore in a follow-up blog; but to my mind premium illusion or premium-aversion is very real, and at play. It alone probably explains the lion’s share of why market participants are so reluctant to actively consider options-based hedges.
As I believe this select example has shown it can be demonstrated that a lack of robust product understanding can really detract from hedging performance.
It simply doesn’t need to.