BACKGROUND

In determining the interest that account holders are paid on cash credit balances and assessed on debit balances, each currency is assigned a reference or benchmark rate, from which a spread is deducted for credit interest and added for debit interest.¹  As account holders may withdraw unencumbered cash balances upon demand and regulations generally restrict the reinvestment of such balances to short-term instruments of high credit quality, benchmarks typically represent the rate at which local banks may borrow on an overnight or short-term basis (e.g., EONIA, Fed Funds).

While the current benchmarks are useful in that they tend to be longstanding, widely accepted and published rates, often used as the basis for determining consumer borrowing, some have characteristics which limit their effectiveness, particularly in the case of brokerage accounts where the spread as applied by IBKR is relatively narrow. A discussion of these limitations is provided in the overview below.

OVERVIEW

Benchmark rates are often determined by either bank survey or actual transactions. The Hong Kong Interbank Offered Rate (HKD HIBOR), for example, is determined by surveying a panel of banks for the rate at which they could borrow funds from other banks of at a specific time each day. The final rate is determined by discarding a set of the top and bottom survey responses and averaging the remainder. Transaction based benchmarks such as EONIA are determined using a weighted average of all overnight unsecured lending transactions by panel banks in the interbank market as reported to the European Central Bank.

There are shortcomings to both methods which, at times, causes them to be an inadequate mechanism for establishing client debit and credit interest rates. Examples of these are provide below:

• Survey rates often represent an offer rate which, by definition stands above the bid rate and can be skewed well above the mid-point when spreads are large;
• Survey rates are typically based upon an inquiry performed at a specific time of the day and may not represent the rates available over a broader period of time;
• The population of institutions surveyed or whose transactions are considered may be small and/or may have borrowing characteristics that are not representative of financial institutions as a whole;
• During periods of market stress, interbank transactions may suffer from reduced liquidity, on either a regional or global basis, thereby distorting benchmark rates.²
• Survey processes often provide little transparency as to how the benchmark was determined and in the past have been subject to manipulation.

AN ALTERNATIVE APPROACH - MARKET IMPLIED RATES

To address these shortcomings, IBKR proposes to implement an alternative method for determining benchmark rates which we refer to as Market Implied Rates. This method combines the optimal attributes of each of the survey and transaction methods and uses as its basis Forex swap prices and the interest rate differentials embedded therein. The Forex swap market is one of the largest and most competitive markets with a daily turnover of 2.4 trillion USD³, representing aggregate transactions well in excess of that used for the current transaction-based benchmarks.

As over 90% of these transactions involve the U.S. Dollar, Forex swap prices of currencies vs. the U.S. Dollar will be sampled over a pre-determined time period referred to as the “Fixing Time Window” that is intended to be representative of liquid hours and primary turnover. The specific swap tenor and fixing windows used depend on the currency. Using the best bid and ask from a group of up to 12 of the largest Forex dealing banks⁴, implied non-USD short-term rates (generally Overnight (T/T+1, Tom Next (T+1/T+2) or Spot Next (T+2/T+3) ) will be calculated. At the Fixing Time Window close, these calculations will be sorted with the lowest and highest disregarded and the remainder averaged to determine the Final Fixing Rate. This Final Fixing Rate will then be used as part of the effective rate for that day’s interest calculations.

To provide complete transparency as to the rates used to determine interest on client credit and debit balances, IBKR has historically posted and updated to the public website each day all of the information an account holder would need to determine the interest they might pay or receive on cash balances (e.g., the stated benchmark, current and historical benchmark levels, spreads and tiers). Similar transparency will be provided with the implementation of Market Implied Rates. Here, rates will be posted to the website in 3 stages:

1. Live – the last benchmark rate calculated prior to the start of the current day’s Fixing Time Window;
2. Fixing Period – represents a running calculation of the current day’s benchmark rate using available data obtained while Fixing Time Window remains open.
3. Fixing – the benchmark rate as calculated upon close of the Fixing Time This rate will remain unchanged for the remainder of the day and serve as the benchmark rate.

NEXT STEPS

Merging interest rate benchmarks and Market Implied Rates is intended to better align the rates offered to clients to the true funding costs and opportunities available to IBKR. The analysis performed thus far suggests that for certain currencies the new benchmark (effective rate) resulting from Forex swap implied rates but capped 25 bps⁵ above/below the benchmark fixing will be higher at various times and for others lower. As for the impact to clients, a higher benchmark generally benefits depositors and a lower, borrowers. What is important is that the new methodology is calculated in a consistent manner, using readily available and substantially representative data.

As the proposed change is significant in terms of its logic and its potential impact to certain clients, IBKR has been calculating and displaying, but not yet applying, market implied rates until clients have had sufficient opportunity to review the data. By August 1, 2017 we will start migrating the benchmarks from fixed to the new system where we use effective rates which are composed of market implied interest rates capped 25 bps above or below the current benchmark fixings. 
 

______________________________________________________________________________________
¹ In the case of the USD, a spread of 0.50% is deducted from the benchmark for purposes of credit interest and a spread of 1.50% added for purposes of debit interest. The benchmark rate for the USD is the Fed Funds Effective Overnight Rate.

² Examples of this were experienced during the financial crisis of 2007-2010.

³ Source: BIS Triennial Central Bank Survey, Forex turnover April 2016. http://www.bis.org/publ/rpfx16fx.pdf

⁴ The actual number of banks selected may vary by currency.

⁵ The 25 basis points is subject to change at any time without advance notice.