FX firms often struggle to provide adequate pricing for exotic currencies during volatile market times. When prices fluctuate sharply, it’s too risky to rely on limited data sources.
Multiple independent and tested feeds – bundled together – are imperative to get an accurate view of the situation.
Exotic currencies are traditionally less liquid because they tend to be poorly represented by liquidity providers. In times of market decline, many stop trading them altogether to avoid risk, reducing liquidity even further.
A composite FX feed can help brokers mitigate risk during market turbulence and provide better pricing to their clients.
Issues with the Turkish Lira and Other Exotic Currencies
The Turkish lira has been among many exotic currencies that have suffered this past year. The lira’s liquidity is unevenly distributed and mostly concentrated in Turkish banks.
Whilst western banks provide all the necessary conditions and capabilities to trade, the unstable market situation has set a bad tone for taking on volatile exotic currencies.
Investment-wise, there’s nothing foolproof about the lira.
In order to have adequate liquidity, a broker would need to gather data from over five of the largest G10 banks to better diversify its providers. Relying solely on traditional liquidity providers isn’t the ideal situation.
Traditional providers generally solve insufficient liquidity issues by connecting brokers that want Turkish lira to local banks with real market involvement. This ensures accurate pricing.
This issue is relevant to practically all other emerging market currencies. The COVID-19 crisis hit hard causing sharp depreciation and deeply depressing demand.
In February, the Mexican peso dropped by over 20%. Mexico, the second-largest Latin American economy, experienced widespread shortages in FX liquidity. The Brazilian real lost over 25% of its value as Brazil faced similar liquidity issues.
Providing accurate pricing during irregular market conditions is not a trivial task. Amid chaotic liquidity offerings for exotic currencies, a solution exists and now brokers can get quality, real-time data feeds.
Aggregating Quality Market Data in One Superior Feed
We’ve developed quality composite Forex feeds with methodological transparency. A composite feed is the weighted average of a set of source OTC FX feeds.
Each average is a sequence of the bid/ask prices sourced by liquidity providers and banks. The source feeds and weights are carefully selected to optimize the following aspects of the resulting feed:
Put another way, the composition aims to average market data from many providers while assigning the weights proportionally to how much a particular source increases the final quality.
The cost of composite feeds depends on the number of demanded symbols with volume discounts.
Cherry-Picking Feeds for the Best Quality Metrics
One of the most welcomed features of an aggregate feed is having frequent updates. No consensus exists among market participants on how to define this precisely, but they often feel uneasy when prices don’t move for a prolonged period of time.
The composite feed minimizes idle time by setting a pre-specified time threshold when no bid/ask updates occur.
Reference feed distance
Reflecting the current FX market situation is a prerequisite for a composite FX feed. To do that, we use a reference midpoint feed with the following properties:
The higher the similarity of any other feed to the reference feed, the higher the quality.
Median spread width
When given two feeds with all other conditions being equal, the one with the tighter spread is preferred. This metric is a common proxy to available liquidity. Highly liquid environments are generally associated with greater credibility.
Here’s what we do to compose an FX feed in the abstract.
First, for a given currency pair, all available market data providers are considered. The three quality metrics are computed on a representative data sample from each provider.
The obtained values are used to rank the providers within each metric. The resulting ranks are then summed. This allows us to efficiently order the available source feeds by the sum of the ranks.
Next, for N source feeds, different N weight sets are estimated: using the best feed (useful for comparison—it is trivially 1), using the first two best feeds, etc.
A weighted average is computed so that it approximates the reference feed as closely as possible.
Thus, we assign smaller weights to the feeds that contribute to increasing the distance between the reference and the composite feeds, and vice versa. Technically, we perform the optimization using a numerical routine and validate our results via extensive backtesting on historical data.
Finally, the three metrics are evaluated on the obtained synthetic feeds and the best one is selected.
According to this method, a composite feed for the Turkish lira can be a reliable source of market data aggregated among the best liquidity providers for the currency, namely local Turkish banks with a real involvement in that market and ordinary, steady liquidity providers.
The composition aims to average the market data from many providers while assigning the weights proportionally to how much a particular source increases the final quality.
We also thoroughly check all source feeds to ensure the resulting composite feed is fair and not augmented by non-existing or implied events in the quote flow.
Brokers looking for real-time access to accurate outlooks on the state of the FX market, especially to its segments with limited liquidity, will significantly benefit from receiving composite FX feeds.