Yield Curves (term structure of interest rates) – filling in the blanks
| 27-05-2016 | Lionel Pavey
Most treasurers do not have access to a dedicated financial data vendor (Bloomberg, Reuters) but are regularly faced with having to discover prices related to yield curves. There are websites that can provide us with relevant data, but these are normally a snapshot and not comprehensive – the data series is incomplete. It is therefore up to the treasurer to complete the series by filling in the blanks.
A quick refresher about the construction of a yield curve raises the following points:-
- All data must be from the same market (treasury bonds, Interest Rate Swaps (IRS) etc.)
- A regular term (maturity) is preferred for ease of construction
- A curve must be smooth
- An implied zero yield curve can be built from the smooth par curve – a theoretical yield curve where no interest is paid until maturity. In a bond this would redeem at par (100) and be issued at a deep discount to par
- A series of discounted cash flow factors (DCF) are produced
- An implied forward curve with constant maturities can be built from the par curve
- An implied forward curve must be monotonic – each point in an increasing sequence is greater than or equal to the preceding point, each point in a decreasing sequence is smaller than or equal to the preceding point
If we look at IRS par yield prices that can be found on a website, we can regularly see yield prices for periods from 1 year to 10 year inclusive, a 15 year price and a 20 year price. To construct a complete curve from 1 year up to and including 20 years we need to fill in the blanks at 11,12,13,14,16,17,18 and 19 years. These yields are assumed to be par yields – the coupon rate is equal to the yield to maturity and the instrument trades at par.
Before starting let us define the procedure for constructing a par yield curve:-
The methodology used is called “bootstrapping”. This allows us to extract discount factors (DCF) from the market rates. DCF’s allow us to calculate a value today for a cash flow in the future.
We assume that the nominal value for all calculation purposes is 100
For a 1 year rate we know the interest and redemption amount at maturity. A DCF is built whereby the net present value (NPV) of these future cash flows in 1 years’ time is equal to 100 or par.
For a 2 year rate we receive interest after 1 year and interest and redemption amount at maturity.
We discount the 1st years’ interest with the DCF we obtained from the 1 year rate and deduct this amount from our initial nominal of 100. This net amount is then divided by the interest and redemption at maturity (at end of 2 years) to obtain the DCF for the 2 year rate.
Example:
1 Year 7% 2 Year 9%
1 Year
100 / (7/100+100) = 0.93457944 (DCF)
2 Year
9 * 0.93457944 = 8.41121496
100 – 8.41121496 = 91.58878504
91.58878504 / (9/100+100) = 0.8402640829
These DCF’s can then be used to find the NPV of any cash flow maturing in 1 or 2 years’ time.
The following example shows a yield curve from February 2013 published on the website of an interbank broker.

yield curve February 2013
The quickest way to price the missing periods would be with straight line interpolation of the par curve between the known points – which would produce the following par curve, zero yield curve and forward curve with constant 1 year maturity.

yield curve February 2013

Straight line interpolation
Whilst the par curve and zero curve are smooth, the implied 1 year constant maturity curve is jagged and certainly neither smooth nor monotonic. The 11th 1 year period rate is lower than the 10th period and the 15th 1 year period rate is higher than the 16th period.
A second approach would be to apply a weighting to the known periods of the par curve and to average the difference out over the missing periods. Read more on this second approach in my next article which will appear next week.

Treasurer



Last week I visited an information session about financial postgraduate education. It was organized by the VU (Vrije Universiteit, Amsterdam). I noticed an increased interest in comparison to last years session, which is great. Information was provided about courses I see back in the CV’s of treasurers: CFA, RBA (Register Belegging Analyst) and of course RT (Register Treasurer) that has an overlap with the ACT courses. Education, specifically postgraduate, is a topic that returns in many of my meetings. This is what I notice on the topic:
Today in Rob Söentken’s Option Tales: When buying options it is tempting to see if the premium expenses can be minimized. A number of solutions are possible, which will be discussed in four articles. In the
bought the option for 1.5%, we could sell it after 3 months at 1.1% and buy the USD through an outright forward transaction. This approach shows that the net cost of option protection would be only 0.4% (1.5% – 1.1%). Which would be cheaper than the premium of a 3-month option with the same Delta. Also, because the option has a higher Delta than a 3-month option with the same strike (25% vs 10%, see diagram 2), it will follow the spot market much better. The bottom line of paragraph 3 is that a longer dated option can be bought with the intention to sell it again at some point, the net cost being less than buying a shorter dated option. While it serves as a hedge against price changes.
buy the 1-year option in diagram 2 for 1.5%. Alternatively we could consider buying a right to buy this option for 0.4% in 3 months time. At that time the 1-year option will only have 9 months remaining, but the strike and 1.5% premium are fixed in the contract. On the expiry date of the compound option we can decide if we want to pay 1.5% for the underlying option. Alternatively, assuming nothing has changed, we could buy a 9 month option in the market for 1.1% (see diagram 2). In such a case we wouldn’t exercise the compound option.








Should we need to exercise the option to get our USD, it still means a combined hedging cost of 3.8%. Which is more than if we had bought the ATM option for 2% premium. Conclusion: Buying an OTM option reduces the up-front cost versus buying an ATM option. But ex-post hedging with an OTM option could result in total hedging cost which are higher than an ATM option.
option is exercised, or the USD must be purchased from the market at the prevailing rate.
Claire graduated as chemical engineer at the TU Delft. After an internship with JP Morgan she decided to pursue a career in the financial sector. She continued as investment banking trainee with ABN Amro / RBS and for almost six years, half of which in London, she worked in M&A and Corporate Finance. Since 2006 Claire was increasingly involved in renewable energy projects in faraway places (Antarctica, Himalayas) and from there on it was a small step to join Fastned and strengthen the team with her financial expertise.
