unboxed.gmbh, fintech

A daily options pipeline gave Unboxed a box spread curve traceable to the day.

Conditions are read off observed prices, beside the swap and Bund curves of the same maturity.

Client
Unboxed
Field
Fintech, market data
Services
Strategy, data engineering, development

Challenge

The rate depends on prices that move every day.

A box spread combines four options into a position that settles at a fixed amount on a known date. The rate it implies is the rate at which money is placed or raised through that position, and it follows the prices quoted on the exchange that day.

Unboxed needed to state a condition to a client and still stand behind it weeks later. That requires each figure to point back to a price observed on a named date, from a named source, with the calculation in between written down.

The comparison carries as much weight as the rate itself. Money placed through a long box spread is read against the German government bond of the same maturity. Money raised through a short box spread is read against the swap rate plus a credit spread.

In numbers

What the data layer produces.

Maturities on the daily box spread curve
1-10 yr
Term range in the rate calculator
12-120 mo
Spread history against the swap curve
5 yr
Versioned snapshot of every source
Daily

Approach

We built the data layer before the calculator.

Automated jobs collect market data at fixed times on every trading day: the exchange option data from which the implied rate of a box spread follows per expiry and strike pair, together with the comparison curves of swap rates, German government bond yields and the rates of the European Central Bank.

Each run writes a complete daily state into a versioned database and leaves the earlier days standing. From the option data we take the median across all strike pairs per expiry and interpolate onto the standard maturities, which gives one curve per trading day for one to ten years.

Values that are missing or fail the plausibility checks are shown as a gap, so a thin trading day stays visible in the series. Every stored value carries its observation date and its source, and every published figure carries its cut-off date.

What we built

Four tools and the database under them.

  • Rate calculator for both directions: money placed through a long box spread, money raised through a short box spread. Nominal amount, term and spread are set by the user. The output is the effective rate, the comparison rate and the difference in basis points and in euros over the whole term.
  • Box spread curve for one to ten years, derived each trading day and placed beside the swap and Bund curves, so a premium or a discount can be read off directly.
  • Spread history against the swap curve and against the German government bond, annotated with the market events that moved it.
  • Monthly series of the cumulative excess return against the German government bond, from the same data.
  • Versioned database under all four, holding one complete snapshot per trading day.

Result

Every figure on the screen names its date and its source.

Unboxed reads a condition off observed prices and can show where the number came from. The calculator, the curve and the history all draw on the same daily state, so two tools opened on the same day answer with the same figure.

The series grow with each trading day, which lets the behaviour of the instrument be read across an interest rate cycle and through periods of stress.

The tools are indicative and describe the instrument. Any wording that touches regulated advice is set by the client and its regulatory partners.

Have a puzzle?

Tell us about the project that standard platforms cannot handle.