(The opinions expressed here are those of the author, a columnist for Reuters.)
By Andy Home
LONDON, Aug 15 (Reuters) - The London Metal Exchange (LME) is now the only open-outcry market in Europe and one of the last in the world.
The London Stock Exchange closed its trading floor as long ago as 1986. The London International Financial Futures Exchange permanently shut the last of its pits in 2000, taking with it the bright-jacketed traders that enlivened the grey Cityscape.
The electronic tide has since swept all before it, leaving the LME one of the last human hold-outs with its ring of red leather seats, baffling hand signals and old-school etiquette.
The LME's nine ring-dealing members were therefore understandably nervous about the exchange's intention in mid-March to trial an electronic close for the last of the ring sessions.
The results of the three-month experiment, however, generated no compelling evidence either for or against open-outcry trading.
The members themselves, meanwhile, seem to have discovered a startlingly simple way of boosting the amount of trade conducted across the dealing floor.
The longevity of open outcry on the LME - owned since 2012 by Hong Kong Exchanges and Clearing - has surprised even the metals-trading community. Few would have bet that the ring sessions, dating back to the exchange's formation in the Jerusalem Coffee House in the 1870s, would have made it into the 21st century, let alone still be going today.
But they perform two critical functions.
The first is to facilitate time-spread trading, a labyrinthine structure on the LME due to the exchange's unique daily prompt date structure.
The second is to generate two globally significant price reference points over the course of the trading day.
The "official" prices set during the lunchtime ring sessions are embedded in physical supply contracts around the world, forming the core of the exchange's price discovery role.
Industrial hedgers are "highly supportive of Ring settlements", to quote the LME's "Strategic Pathway" document from 2017.
However, the physical supply chain has little or no interest in the second pricing point - the "closing" prices that are set at the end of the afternoon ring sessions.
Financial players, on the other hand, have no interest in the "official" prices but do care a lot about the "closing" prices because they have mark-to-market and margining impact.
This part of the LME user community would also "in general prefer an electronically-derived closing price," according to the exchange.
Hence the idea of a trial for one metal, nickel, using the exchange's electronic platform to generate a volume-weighted average price (VWAP) "close". The experiment was conducted only on "outright" three-month metal with spreads trading staying on the floor.
Complaints about the accuracy of the LME "closes" have been around as long as anyone can remember, even though past bad behaviour such as "banging the close" has largely been remedied by successive waves of regulation.
The electronic nickel trial has revealed some divergence between ring-determined and VWAP-set closing prices.
The LME noted that 40% of ring-discovered closing nickel prices were outside of the available price range on the LME electronic Select system during the one-minute slot used to determine the close.
However, the percentage fell to just 8% over the two minutes following the close, suggesting price divergence is fleeting as human and electronic systems align themselves.
The momentary disconnect may also, the LME conceded, be due to the nature of ring-trading with traders only committing themselves to a bid or offer at the last possible moment.
On average the first quote during the one-minute closing session for three-month nickel came in just 20 seconds before the end of the session, generating a cascade effect that rolls into the minutes just after the session.
As to whether electronic trading of the close leads to more volumes, the results were decidedly ambivalent.
Total average daily volumes rose 6% with greater direct participation in the closing price process but with a drop-off in indirect flows such as "trade at close" orders.
What the LME calls "informal feedback" suggests no consensus on whether the trial made it easier or more difficult for brokers to meet customer demand for a guaranteed closing price.
The lack of compelling evidence either in favour of or against electronic closes is probably why the LME has received no formal feedback at all. It has just extended the deadline to the end of September and detailed some of its findings to encourage more responses.
A separate experiment with the closing price process on zinc, meanwhile, has generated a far more clear-cut result.
During the trial, also conducted from mid-March to mid-June, the closing zinc session was extended from 5 to 10 minutes. The four minutes reserved for spread trading were extended to eight minutes and the one minute session used to determine the closing price of three-month metal was doubled to two minutes.
That doesn't sound like much of a change but resulted in a 57% increase in average daily volumes for business transacted in the closing zinc session.
The extension "helped dealers in the Ring establish bids and offers in a more structured manner"; resulted in "fewer unwanted positions" on dealers' books; and generally facilitated interaction between open outcry and electronic trading systems, the LME said.
There was no negative feedback and the LME "understands the trial has been successful and has had the desired effects for all market participants."
The extension on zinc will become permanent and will be expanded to include all the LME's other core base metals contracts effective from Sept. 23.
Given that the LME's open-outcry system has been battling for years against a dwindling number of ring-dealing members and the continuous erosion of liquidity in favour of electronic trading, it's surprising that no-one thought of such a simple solution before.
The jump in volumes tells its own story but the tangible improvements to market structure and order execution may also help address some of the pricing discrepancies around the close revealed in the nickel trial.
It's a clear win for open outcry trading. The electronic nickel trial, by contrast, appears to be more of a messy draw.
After 142 years it looks like the LME's ring sessions have a lot more life in them yet.
Editing by Susan Fenton