Portfolio/Basket Trading

In recent years an increasing amount of trade volume has been transacted in the form of portfolio or program trading. The introduction of electronic order-driven trading and improved market transparency has lent itself to the simultaneous execution of baskets of stock far more cost-effectively than was possible previously. Most investment banks and major market makers have dedicated program trading desks that will handle the execution and risk of such transactions.

Program trading generally falls into two main types:

  1. 1. principal transaction: where the bank will quote a price and take on the risk of the execution; and
  2. 2. agency transaction: where the bank acts as agent and executes on behalf of the client on a best endeavours basis.

The major players in London who are prepared to devote substantial amounts of risk capital to program trading are:

Programs can be carried out for a number of reasons and are often cheaper for the institution to execute than trading the stocks individually. A transitional trade where one fund manager has lost an investment mandate to another will often be structured as a program, as the new fund manager will take the opportunity to re-weight the portfolio as he sees best. Many programs are conducted on a pan-European level with the program desk utilising the index future in individual markets to hedge net market exposure.

Principal portfolio trading

A client, typically an institution with a basket of stocks to buy or sell, will approach several investment banks for a quote.

Only general portfolio characteristics are given at first

For instance, the institution may have a list of twenty-five stocks that it wishes to buy and twenty-five stocks that it wishes to sell, perhaps to re-weight a portfolio. Rather than execute every stock separately, the fund manager may ask the investment bank to quote on risk. Only basic characteristics of the trade are revealed at this stage, for instance the size of the long positions to sell and securities to buy (for example £120m long to sell and £100m new positions to buy), percentages of average daily volume, NMS multiples, correlation with an index, etc. The portfolio desk will then respond shortly with a two-way quote expressed in basis points around middle prices and the institution then usually awards the business to the most competitive desk.

The winning firm is then provided with the details of the portfolio, securities and number of shares to be transacted a few minutes before the agreed transaction time. Pre-hedging, where the firm starts to cover its risk on the transaction early, as soon as it is aware of the details of the trade, is legal so long as the client is informed that it intends to do so. The client has to balance the risk of giving specific constituent information to a firm that may not eventually win the mandate, against receiving a less competitive quote for more general portfolio characteristics.

The prices, usually middle prices, are then frozen at an agreed time, avoiding times when key figures such as non-farm payrolls and interest rate announcements are released. To protect the firm from spoiling tactics, the trades are usually reported at the prevailing middle prices and the firm’s risk premium is taken in the commission agreed.

Often portfolio trades will be executed on the market open, or at-market close, or on the hour during the trading day at the request of the client. These simultaneous multiple trades can account for some of the odd looking at-market orders in the opening or closing auctions, or for strange short-term moves in individual stocks during the trading day.

Aggressive portfolio desks

Periodically, to win business, a bank may take a very aggressive stance in bidding for portfolio trades, even to the point where it is not commercially viable, in an effort to buy market share. Trades will often be priced based on the firm’s relationship with the client, although there may sometimes be a bigger picture and the deal may be aggressively priced to buy order flow. Bigger firms with large amounts of capital available to deploy and big flows with an opportunity to match business, may be better placed than smaller ones. A strong presence in program trading can create a virtuous circle with increased flows to the cash equities desks, improving execution and pricing. The firm may hope that the successful execution of a portfolio trade for a content client may win business for other parts of the firm, perhaps in cash equities, block deals or corporate finance. A presence in portfolio trading also makes a clear statement of commitment in the business. However, direct access to the cash market and crossing networks may, in the future, pull some liquidity away from the portfolio trading desks.

Portfolio trades are easy to spot and the trade codes usually used for principal transactions are ‘N’ and ‘P’. There have to be at least twenty stocks for the trade to be classified as a portfolio trade by the London Stock Exchange.

Basket trade is now a generic term but strictly speaking a basket relates to a group of stocks delivered in proportion to their weighting in a particular index or benchmark.

N non-protected portfolio

N trades will quite often appear across several stocks simultaneously as the basket is reported within three minutes of trading. The prices will have been struck at an agreed time and the trades automatically have a publication delay of:

  • • 120 minutes for SETS stocks; and
  • • 60 minutes for SEAQ stocks with an NMS of greater than 2,000.

The prices will sometimes seem odd and may run to two decimal points or more as, for instance, the middle prices will be frozen and an offset applied. Sometimes the price is struck at middle prices and the bank’s profit (risk premium) will be incorporated in the commission charged (e.g. 6-8bps).

P protected portfolio

Unlike a non-protected portfolio, the market maker benefits from a delayed reporting of the trades, normally until the end of the day or when the risk is 100% offset. Similar to the protection in an individual stock, the institution is guaranteed a certain minimum price for the portfolio, but, if the execution goes well during the day, may receive an improvement in prices.

Protected means that there is an intention to improve on the transaction within a specified period. Of course this is at the discretion of the market maker and his relationship with the institution. A flurry of P trades will often go through the market in the last fifteen minutes or so of trading, as the orders that have been worked are trade reported and published immediately. Reporting the trade as a protected trade allows the firm to benefit from not having to report the transaction until the end of the day when publication will be instantaneous.

Again, the improvement in price may be reflected in a reduction in commission charged rather than an adjustment to the actual reported prices.

Anything between 0-5 basis point’s slippage is regarded as reasonable execution. It is not in dispute that the introduction of electronic trading has reduced bid-offer spreads, as the London Stock Exchange is always keen to publicise, although it is not a full picture without knowing what size is on the bid and offer! It may only be ten shares. However, average spreads for FTSE 100 stocks are now less than 15bps, with FTSE 250 typically less than 65bps.

Agency portfolio trading

Here the firm executes the portfolio on a best endeavours basis and the client takes the execution risk. The client may give the firm some criteria to execute against such as:

For example, if the trade consisted of selling £40m of stock to buy £30m, the client may ask the firm to raise £10m immediately by selling 25% of the portfolio then transact on a money for money basis.

Transaction-based management systems allow the client to monitor progress of the order electronically as it is being executed and he can adjust his instructions accordingly. If some stocks have been difficult to execute against a VWAP target for instance and suffered from slippage, the firm may smooth the prices by adjusting the better-executed transactions to allow a better fill for the worse ones. In other words, if some prices were more than 10bps away from VWAP, both better and worse, the prices of the worst trades would be adjusted to make the spread tighter with an overall range of say +/- 3 basis points.

If a firm guarantees execution against VWAP then any improvement in execution goes in the firm’s pocket, but the firm also takes the risk of slippage. Agency portfolio trades generally attract commission levels of less than 10bps, but there seems only one way that commission levels are headed.

One contentious area is chopping, where an unexpected improvement in execution is retained by the firm instead of being passed onto the client; however, modern online order management by the client in real time makes the process more transparent and the client can always request full time and sales details if suspicious.

Risk of execution against VWAP

Guaranteed execution against VWAP carries some risks. Suppose, for example, a desk is asked to sell 10m Shell shares during a day’s trading and to match VWAP, and the average daily volume in Shell is 100m shares. By lunchtime the firm has executed half the trade and, as expected, 50m Shell has traded so far, with the price staying stable at 375p. Some unexpected news comes out and the stock rallies dramatically to 390p, where heavy trading takes place, and at the end of the day 150m shares actually trade, with the day’s VWAP somewhere near 385p. The broker now has no chance of matching VWAP as most of the volume has taken place at the higher price. Also, because the day’s volume was bigger than expected, he sold too much too soon.

The booking of portfolio trades at the end of the day represents the aggregated total of many transactions that have taken place during the day and can often explain sharp stock moves earlier in the day. Many novice traders do not realise that simply watching buys and sells going through the market is not an indication of the whole picture, due to the delays in trade reporting and publication enjoyed by the providers of risk capital.

It is a feature of the European, and in particular the UK markets that a substantial amount of price-sensitive information is released throughout the day, unlike the US, where most announcements are made pre- or post-market. This policy of not suspending stocks gives a big advantage to direct access traders in the UK, but can be tricky for dealers working an order against VWAP, or as part of a program trade, and can be a big shock for US program trade experts who are brought over to Europe, having previously only traded the US domestic market.

Futures and index arbitrage

The FTSE 100 Index futures contract is the most important contract as far as a UK trader is concerned, although it will be influenced by movements in other European indices and, of course, the major US indices. The FTSE 100 future offers a cheap means of speculating on the future direction of the UK market, hedging a position, or arbitraging between the future and the underlying cash indices. A detailed explanation of futures contracts is beyond the scope of this book; however, I do believe it is important to understand how the futures contract can be utilised, what its relationship with the underlying cash market is and how that can affect moves in the underlying cash market.

FTSE 100 Index

Trading the FTSE 100 contract allows market participants to gain exposure to the main UK blue-chip benchmark index through a single contract. The FTSE 100 is a capitalisation-weighted index of the top 100 (actually 102, as two companies, Schroders and Royal Dutch Shell have two classes of share) blue-chip UK companies, representing 80% of the UK market by value. Capitalisation-weighted means that a 5% move in a large company such as BP will have a bigger effect than a 5% move in a smaller company such as Sage Group. There is also a futures contract on the FTSE 250 but that is relatively illiquid.

FTSE 100 futures contract

The FTSE 100 futures and options contracts are cash settled at the Exchange Delivery Settlement Price (EDSP). This means that if you buy a futures contract and hold it until expiry, you do not take delivery of the underlying basket of stocks, but receive or pay a cash figure depending on the difference between the opening price and EDSP.

Until November 2004, the EDSP was based on the average value of the FTSE 100 Cash Index every fifteen seconds between 10.10 and 10.30 on expiry day, which is always the third Friday of the delivery month. Of the eighty-one measured values, the highest twelve and lowest twelve were discarded and the remaining fifty-seven averaged to calculate the EDSP.

In November 2004, the London Stock Exchange introduced a new procedure whereby the final EDSP was calculated using an intra-day auction in the underlying stocks, using the same concept as used in the closing auction except that there is no volume check. Expiry now starts at 10.10 with the uncrossing taking place at 10.15, plus a random start time. This new procedure was introduced to concentrate expiry liquidity into a shorter period of time to avoid the wild swings that have occurred in the index during previous expiries.

Note: SETS random periods are a maximum of thirty seconds, whilst SETSmm are a maximum of sixty seconds.

Futures contracts trade on a quarterly cycle of March, June, September and December so there is an expiry every three months. The most liquid contract is usually the closest to expiry – the front month – until very close to expiry when the next contract out will become the more actively traded. Trading hours are 08.00 to 17.30.

Fair value

The key feature of futures contracts is the concept of fair value. Normally the futures contract trades at a premium to the underlying index, even though the two are guaranteed to converge at EDSP at contract expiry. This is because if you were to buy a basket of stocks now, you would need to finance the position until expiry date, so it is more cost-effective to buy the futures contract. This is reflected in the premium that the contract trades at. However, stock indices contain companies that pay dividends, and if the stock goes ex-dividend between now and the expiry date you receive that dividend by holding the basket of stocks but not if you hold the futures contract. Therefore the fair value of a futures contract can be calculated based on current interest rates and expected dividends.

Say the FTSE 100 is trading at 5000 and there are 80 days until expiry, with interest rates at 5%. The cost of financing a long basket of FTSE 100 stocks will be:

5000 x 80/365 x 5/100 = 54.8 index points

If you have also calculated that there will be expected dividends worth 22 index points, then the fair value for the futures contract will be:

5000 + 54.8 - 22 = 5032.8

So if the futures were trading at fair value, with the cash Index at 5000, you would expect the future to trade at 5033.

Arbitrage keeps the futures close to fair value

Of course, the futures contract will fluctuate with supply and demand; however, if it rises too far arbitrageurs will step in and sell the futures contract to buy the underlying stocks, hence pushing the cash market up. Similarly, if the futures contract falls, they will buy the contract and short the underlying stocks. Due to the cost of executing transactions in the underlying market, such as the bid-offer spread, there will be an arbitrage channel, within which it is not worth executing the trade.

Take an underlying cash Index value of 5100, with a bid-offer spread of say 0.2% (equivalent to approximately ten index points); the bid price of the index constituents all added together might be around 5095 and the offer price 5105. If the fair value of the futures contract were, say, 55 index points, and an arbitrageur wanted to lock in a profit of at least 20 index points, he might not start selling the futures contract until it reached:

5105 + 55 + 20 = 5180

Possible risks with arbitrage

Contrary to popular belief, index arbitrage is not risk-free. A couple of interesting situations can occur:

  1. 1. If you are running a long stock basket/short futures arbitrage position and a stock is suspended and removed from the index at a certain price and replaced with another, because the stock is removed at the suspended price you take the hit if it subsequently turns out to be worthless. This is because the index has not fallen, as it would have done if the stock had been removed at zero.
  2. 2. Cash flow can be an issue. If you are short a large futures position in a rising market, you will be required to post additional variation margin with the futures exchange; similarly with a long futures position in a falling market. This can lead to serious cash outflows and although you are fully hedged you will receive no credit for the hedge position in the cash market.

There is also the execution risk involved in unwinding the transaction on expiry day, covered in a later chapter.

Fair value is not the same for everyone

Fair value is a subjective calculation and may vary between firms, depending on their internal financing costs and assumptions on forward dividends. In addition, different firms may value the dividends differently based on their own tax position. Firms with large amounts of capital and flow in the blue-chip stocks will be particularly well placed and there will be a close interaction between the cash desks, the program trading desks, and the index arbitrage desks.

The multiplier on the UK FTSE 100 futures contract is £10 per point so buying one futures contract at 4550 gives effective exposure to £45,500 of FTSE 100 stocks. The Exchange has the right to raise margin rates at times of high volatility but usually the margin rate is around £1,500 per contract, which, with an index level of 4500, corresponds to around 3% or 30:1 gearing.

Index arbitrage no longer a monopoly of major market makers

Until the introduction of electronic trading in 1997, index arbitrage was the franchise of the main market makers as they had a monopoly on making prices, the ability to borrow stock to short sell and to trade without paying stamp duty. Now automated index arbitrage trading strategies allow computers to automatically execute optimised baskets against the FTSE futures contract.

As an aside, it is interesting to note for technical traders that the charts of the underlying FTSE 100 cash Index have changed demonstrably since 1997. This is because the Index was calculated by taking the middle price of the competing market makers quotes in all stocks pre-1997 and often there would be a gap up or down in the opening Index price when market makers reacted to significant overnight news. In other words the Index was being calculated based on simultaneous quotations rather than actual trades. Now the Index is calculated on the last traded price for each stock and, as it may take some time until all stocks are traded after the open, the cash Index tends to creep up rather than gap up because with electronic trading the stocks come up in turn.

Exchange for physical (EFP)

An institution may approach an investment bank to transact an exchange for physical transaction or vice versa. A simple example might be where a fund manager has bought 500 FTSE 100 futures contracts as a quick and cost-effective way of gaining market exposure and now wants to convert into the underlying stocks. If the fund manager wanted the exact constituent stocks of the underlying index, this would be a very simple transaction and might even suit the existing book of the investment bank. The arbitrage desk may have a short futures long stock position established for a differential of, say, 68 points with fair value currently 52 points. Running this to expiry will guarantee a profit of 16 index points, after execution slippage. Here, however, the bank may offer the exchange to the fund manager at, say, fair value or slightly below. The 500 contracts are crossed in the futures market at the prevailing market rate, so the fund manager closes his long futures position and the investment bank closes its short, and the investment bank delivers the basket of stocks to the fund at prices representing an effective index value of 52 index points below the level at which the futures were crossed.

This is a simple example. What is more likely to happen is that the fund manager wants a specific basket of stocks that may be similar to the underlying basket of the futures contract but not identical. The bank will then price the trade according to the residual risk involved in handling the skewed weightings.

The futures and cash markets

It is overly simplistic to define any one trading day as futures-driven or cash-driven, but because of the strong interaction between the two markets, linked by arbitrage activity, it is important if trading one to keep a close watch on the other. If an investment bank takes on a large program trade in the cash market that is heavily weighted to a net large long or short position, it will almost certainly seek to hedge some of that exposure in the futures market. Institutions seeking quick exposure to UK equities can buy futures quickly and inexpensively, and then later switch these into a basket of stocks.

Quite often, therefore, the balance of the day’s action will be in one or other of the markets and one will lead the other. It is essential for the UK trader to have a live FTSE 100 futures feed as well as to keep a close eye on the other main European index futures markets and the early show for the US futures markets. The Dow Jones, S&P 500 and NASDAQ futures all trade pre-market during UK morning hours, giving an early indication to the open of US trading.

Equity block trades and bought deals

The UK market has always embraced a principal, risk-taking culture, although the process is now far more refined than it was. The two big risk-taking houses, Smith New Court and Warburg, dominated the late eighties and early nineties at the time, particularly the former, whose placing power was second to none.

Risk prices

If an institution or shareholder with a chunky stake in a company wanted to sell that stake in one transaction, rather than drip feed it onto the market and depress the share price, they would approach one of the investment houses for a risk price. The broking firm would then bid the shareholder a price at a discount to the prevailing market price of, say, 5-10% on risk, depending on the liquidity of the stock, market sentiment at the time, and the bank’s perception of the risk. If the price was accepted, it was then up to the bank to go to work and place that stock as cleanly, quickly, discreetly and efficiently as possible, before the market place got wind of the placing and the share price faded. This was much easier to do before electronic trading arrived, as all stocks were quoted on SEAQ, and a market maker could sit on the bid and make the stock look better than it actually was.

For simplicity’s sake take, for example, a stock which is quoted at 545-550p. The investment house may have agreed to buy the stock at 521p and place it at 525p, simultaneously making a turn of 4p per share. It doesn’t sound much, but when you are placing 50m shares, that adds up to £2m on a £260m order. The house conducting the business would invariably sit on the bid and hold the stock up for as long as possible, to make the deal look as attractive to institutions as possible. It is one of those rare situations where a stock can be offered in big size at 525p but be bid for 545p in the normal market size. The booking of the trade would obviously take place later or at the end of the day, giving the house some protection from spoiling tactics.

Nowadays the business of bought deals is far more sophisticated, and the monopoly that was enjoyed by the big two market makers has been relinquished to the main global investment houses like Goldman Sachs, Morgan Stanley and Deutsche Bank.

Book building exercises

A bank will often use a book building exercise to gauge market sentiment, rather than attempting to estimate the right risk price itself by offering the shares at a fixed price. Although the risk, of course, is that like any big deal, everyone wants to buy the last 500,000 shares when it is clear the placing will be a success. The investment bank will also try to avoid overnight risk; you can only get a certain amount of protection from index futures hedging, and correlated indices and sectors offer little protection.

Backstops are commonly used, where the bank guarantees a minimum price to the seller and takes the risk of the deal on its own book, absorbing any loss. If a price improvement is achieved the difference is shared between client and bank or a fee paid.

Corporate relationships are critical and bought deals will often be part of a bigger picture bringing pressure to bear on winning the mandate. In other words, the bank may be prepared to accept a thinner margin than the risk of the deal implies for the sake of maintaining a good relationship with the client.

Pressure on banks to take on large trades

Contrary to popular belief, taking on bought deals and large program trades for clients is no less risky than proprietary trading. A proprietary trader always has the choice of whether to take on an opening position, whereas the firm may be under client pressure to put its own capital at risk and take on a large trade initiated by the client. A reputation for taking on large trades increases the probability that the firm will have an opportunity of seeing the next trade. With pressure on banks to maintain their global ranking in equity market league tables, the business has become much more about balance sheet strength than placing power. Prevailing market sentiment on the actual day will always play a big part in the success, or otherwise, of a placing.

The market is so competitive and keenly priced, that sometimes the investment bank will misjudge the risk and appetite for the stock and be forced to take a lower price to get the stock off its books. Both Goldman Sachs and Morgan Stanley took hits in bought deals that they bid for in late 2003; the former while placing Texas Pacific’s 21.8% stake in Punch Taverns; the latter in placing the Italian government’s stake in energy group Enel.

The effect of a bought deal or placing on the market rather depends on how it is handled and whether it leaks out. It is much easier with an electronic market (Central Counterparty preserves anonymity) to spoil someone else’s deal, by short-selling stock aggressively, whereas previously the telephone had to be used, and the market maker had much more control over the quote on the screen.

Clients will be mindful of share price performance leading up to the execution of the trade to gauge the real discount to the market price that they are accepting. If a tight lid is not kept on the transaction, share prices can magically weaken ahead of the placing.

Booking of large trades

The booking out of the trade will actually take place later in the day, and the single large transaction that goes through the trade ticket will represent the aggregated opposite side to the many smaller transactions going through earlier.

Large trades that are booked out and appear on the ticker at prices significantly different from the prevailing market price are likely to have a significant instant effect on the share price.

Example: Geest

Geest shares shot up immediately on 28 May 2004, when 3.4m shares went through the ticker at 580p printed at 15.26 in the afternoon. It was later revealed to be a stake taken by the Icelandic group Bakkavör.

Figure 1.24: Geest reacts to large trade at a premium to prevailing price

Used with permission from Bloomberg L.P.

Reverse bought deals

Reverse bought deals happen rarely but are utilised when a company makes a dawn raid on another company, bidding for a chunk of stock at a premium to the prevailing market price.

A frisson of excitement goes through any dealing room whenever any large program trade, or bought deal, is taken on by a member firm on risk; followed by the elation of success accompanying the Head of Sales Trading shouting the magic words “offer closed” when the deal is successfully completed.

Hanson’s raid on ICI

One of the most famous cases was Hanson’s raid on ICI in 1991, when Smith New Court raided the market on Hanson’s behalf. The deal was taken on risk, in other words Smith undertook to deliver the stock to Hanson at a fixed price, rather than as a less demanding agency trade, in keeping with its risk-taking culture. Smith bought stock quickly and efficiently in the market.

In those days, Smith was a quoted company trading on a sliver of capital, often taking on large bought deals and program trades, brokered by veteran jobber Michael Marks, many times the size of its market capitalisation.

Its cause in this instance though was helped by not revealing the name of its client, and some institutions were furious when the declaration was made later that evening, just before the news services closed, that it was Hanson that had bought the stake.

I was fortunate enough to observe the whole process first-hand, and some days later I looked across the dealing room and caught sight of Hanson gazing silently across the feverish dealing room. After taking a phone call I glanced back again, but he had disappeared as silently as he had appeared. In some ways almost reflective of his silent and unoticed movements through the world’s capital markets.