Creating A Heirarchy Of Trading Decisions

Jot_Wagner.qxp

1/4/06

5:58 PM

Page 1

Creating A Hierarchy of Trading Decisions WAYNE H. WAGNER

I

requirements. The appearance of a defensible procedure for assuring best execution is not the same as achieving best execution. Indeed, best execution might just be the piece left in the box when this decision tree puzzle is completed. The decision tree below performs some perhaps appealing categorizations and divisions, but it misses something crucial. Whatever insights come through from this exercise, it is only a shallow mimicry of what traders do intuitively and better than described here. Not everything that counts can be counted. In trying to codify the activity, we miss the trader’s edge: the unconscious sense of heft and intuitive feel for the market, a conundrum that arises through talent and the experience of having experiences. Excellence comes down to knowing from within, not from without. What was it like to have been there and done that? You can’t read a book on skiing and believe that is sufficient for you to become the next Johnny Moseley. Enough preamble; let’s get started. Without posting some guideposts along the way, we run the risk of drawing a picture that would make a flow diagram of an oil refinery seem intuitive. Therefore, in the sections that follow, the branches in the tree are always identified using italics. In addition, the appendix lays out the structure in tabular form.

LY

n today’s markets, institutional-level trading can be accomplished in an intimidating number of ways. Any one broker or a combination of hundreds may be able to execute the trade, using dozens of possible trading venues and techniques. How does a trader thread through the myriad of options? This article strives to identify a hierarchy of the decisions a trader must make to complete an execution. Please, do not take what follows as a unique insight. This article lays down one possible structural view of trading. Yet most of what traders do is largely tactical; by its very nature, trading is one of the most adaptive of human behaviors. Similar to most interesting human endeavors, trading is not cause and effect but action and reaction. Expert performance of a delicate and interactive task cannot be reduced to mechanical actions performed by a very smart computer. Traders who act with intention, meaning, and value will always trump those who blindly follow a cookbook procedure. The goal of this article is to provoke thinking, not to supersede it. Follow this procedure unthinkingly and any proprietary desk will soon figure out how to anticipate your actions to their profit, all the while offering assurances that you are the smartest trader they ever met. Just as anyone who mechanically trades according to this decision tree would be foolish, so would a compliance officer who believes that any prespecified procedure will define, circumscribe, and protect best execution

FO

R

R

EV

IE

W

O

is Chairman of Plexus Group, Inc. [email protected]

N

WAYNE H. WAGNER

WINTER 2006

STEP 1.0: RECEIPT OF THE ORDER

Orders do not come from heaven; they come from portfolio managers. In some organizations, that is not a meaningful THE JOURNAL OF TRADING

1

Jot_Wagner.qxp

1/4/06

5:58 PM

Page 2

distinction. The key point is that the trader receives marching orders; he does not initiate the activity. The first step in any well-run shop is to double check the order’s validity. Do we own the shares we are trying to sell? Will we have the cash at settlement to pay for the securities being purchased? Do we have authority to trade without client approval? Are there any accounts that won’t permit the acquisition of Acme Liquor, Guns, Oil Exploitation, and Environmental Havoc, Inc.? If these conditions exist, they will need to be addressed before proceeding. Once these hurdles are cleared, the trader makes his first critical decision: how much discretion is available to select the broker and venue for execution?

EXHIBIT 1 Buy/Sell N-thousand shares of XYZ stock

Unrestricted

Restricted

EXHIBIT 2 Unrestricted order

Easy

Liquidity Providing

N

Tough

W

O

mination. In all situations, but particularly the last one, the trader needs to remain observant of changing market conditions, which provides the opportunity to discuss an important point: the difference between trading strategy and trading tactics. I can’t remember the exact quote, and Google isn’t helping me here, but battle plans are only good until the first shot is fired. The same axiom holds with trading— except in the most trivial situations. Trading conditions change as new participants enter and leave the market. Importantly, the trader needs to be sure to include himself as one of those participants interacting in the market. The point is that the strategy is always tentative, and the tactics must adapt to any changing situation. Although the tough trades are far more interesting to think about, we’ll dispose of the easy trades—and the very easy liquidity-providing trades—first. In most easy cases, the situation will not change much, and the trader has something to offer to the market—the absorption of liquidity in exchange for a reasonable commission and impact. The question is how best to make the most of that opportunity.

EV

IE

Trades that are subject to client direction restrictions, tight WRAP fee arrangements, or commission recapture agreements take routing discretion away from the trader. Effectively, the broker, not the trader, owns the right to trade. The trader often has extremely limited ability to enforce discipline on the broker. The only issues that need to be dealt with for these restricted orders are those of sequencing: if this order must be segregated from other orders, should it be executed early or late in the sequence? (See Exhibit 1.)

LY

STEP 2.0: RESTRICTED ORDERS

R

R

STEP 3.0: UNRESTRICTED ORDERS: TOUGH OR EASY?

FO

Now we begin to approach the interesting issues: is this unrestricted order likely to be a tough or easy trade to execute? Only one thing makes a trade easy: liquidity. Liquidity means that the execution of this trade can be readily handled within the normal ebb and flow of buyers and sellers in the market. Some examples of what would normally be an easy trade are ● ●

●

●

any trade less than ten percent of the volume a few thousand shares of virtually any stock of institutional interest perhaps half a million or more shares of a very liquid name such as Lucent, GE, Microsoft, or Cisco buying a security under strong selling pressure or, vice versa, selling a security to numerous and anxious buyers (liquidity providing) (see Exhibit 2)

To assess current liquidity, the trader would call up the symbol on the trading screens and make a quick deter2

CREATING A HIERARCHY OF TRADING DECISIONS

STEP 3.1: LIQUIDITY-PROVIDING TRADES

If all trades fell into this category, we wouldn’t have much need for skilled traders. Here, the other side of the trade is demanding liquidity, and our trade is providing the scarce goods into the market. The cards are in the trader’s hands, and he can negotiate for a good price. The WINTER 2006

Jot_Wagner.qxp

1/4/06

5:58 PM

Page 3

issue becomes one of timing: slowing down will be a good strategy if the other side keeps up the pressure, since later prices are likely to be better prices. Here a trader should be aware of all venues but will probably get better prices by working soft or hard limit orders or indications of interest through attentive but unhurried brokers.

EXHIBIT 3 Easy trades

Not softable

STEP 3.2a: EASY: SOFT DOLLARS CANDIDATES?

EXHIBIT 4

LY

Not softable

Not crossable

O

N

Crossable

W

However one thinks about soft dollars, they remain an important yet diminishing factor in today’s markets. Soft dollars trades fall under the easy branch of the hierarchy because soft dollar brokers generally don’t have the beefed-up trading commitment to handle the hard trades. (See Exhibit 3.) The trader may choose to go around this resource issue by requesting a full service broker to “step out” part of the commission to a soft dollar broker, but most trade desks today have enough easy trades to satisfy the diminishing demand for soft dollars. So, typically, the soft dollar broker would be called up with the order.

Softable

STEP 3.3: TOUGH TRADES: DIFFICULT TRADING CONDITIONS

IE

STEP 3.2b: EASY: CROSSABLE OR NOT?

FO

R

R

EV

Now we come to the branch in the tree where we must deal with the more interesting case of not softable trades. The decision here is whether the trader thinks it worthwhile to try low-cost alternatives such as crossing networks. (See Exhibit 4.) Crossing networks such as Posit and Liquidnet are attractive because they cut out the middleman in situations where he is not needed. Those crossable situations are created where both buyer and seller can be summoned at the same time to execute a direct trade between themselves. The advantage is low cost; the disadvantage occurs when no cross is available. Unfortunately, most crossing networks have low hit rates, so it’s worth a try but not worth trying over and over again when there is no other side. Then the trades must be switched to the non-crossable branch and are likely to be sent to automated markets such as DOT, SOES, automated execution venues, and algorithmic trading engines or lumped into a program trade with similar trades. The virtues of these simple execution venues are that they provide “good enough,” cheap, fast, and error-free execution. Minimal amounts of valuable trader time and attention are expended on situations where skill is unlikely to make a significant difference. Now it’s time to turn from the easy trades to the tough trades. WINTER 2006

There are three major situations where trading can be considered tough: 1. where the trade is large relative to the average volume in the stock; a quarter of a day’s volume is an oftenused rule of thumb 2. momentum: where the stock is moving up in front of a buy order or down in front of a sell order 3. urgent trades, as demanded by the portfolio manager who has identified a trading need that must be filled as quickly as possible. This is often the worst of all possible trading situations a trader is can encounter. It combines all tough characteristics into large, urgent, and momentum trading conditions. An example would be when bad news has come out on a current holding and the manager expects the bad news to compound. He wants out, and he wants out fast to avoid major damage to performance—complete surrender, in effect, to the forces of the market (see Exhibit 5) Institutional-sized trades seldom execute in the same moment they are received. Measured in dollars, current institutional trading data show that 92% of desk orders THE JOURNAL OF TRADING

3

Jot_Wagner.qxp

1/4/06

5:58 PM

Page 4

EXHIBIT 5

EXHIBIT 6 Tough Trades

Large, Momentum & Urgent

Not crossable

chase is on; the stock price is moving away; the trader doesn’t have the luxury of time working in his favor. Here the cost of not trading, or not trading aggressively, may be more than the cost of trading with immediacy and high impact. Waiting might result in ultimate trade prices far worse than could have been attained by stepping up quickly and aggressively. The trader needs to carefully weigh the readily perceived impact costs against the more egregious delay and opportunity costs. Again, the trader would likely call up the heavy-hitting brokerage firms: effective principal trading desks and brokers working hard to find the other side of the trade.

STEP 3.3a: TOUGH LARGE TRADES

FO

R

R

EV

IE

Large trades, even in neutral markets, can overwhelm the liquidity available in the market. The key is to have access to all visible and reserve liquidity on all trading venues. Where that liquidity does not currently exist, it will be necessary either to wait for it or to coax it out. Here we have a situation similar to the non-softable easy trades discussed above, except we would expect traders to immediately check all possible liquidity sources, with uncrossable trades shifting over to more interactive trading as required. (See Exhibit 6.) The trader in these situations needs to be versed in all trading techniques and venues. At any given time, direct market access (DMA), third market brokers, or displayed limit orders often play significant roles. Typically, the best way to coax out liquidity is through the dealers and full service brokers who maintain expensive personnel, facilities, and information networks for the purpose. This part of the business is often described as one of relationships. A trader cannot expect to call up these elaborate facilities once in a blue moon and expect on-demand first-rate service. A buy-side organization needs to remain an important customer to these brokers, and good trade desks are very sensitive to maintaining some clout with key brokers.

N

LY

exceed 10,000 shares, and 25% of ordered shares are not traded on the day that they were transmitted to the trade desk. While these large orders are being worked into the market, conditions often change, reflecting either the order’s motive (news) or the client’s own impact (the Heisenberg effect).

Crossable

O

Momentum

W

Large

Large, tough trades

STEP 3.3b: TOUGH: MOMENTUM TRADES

The difference between momentum trades and large trading lies in the timing dimension. Momentum situations occur when the manager buying a security is joining other managers and traders responding to similar stimuli. The 4

CREATING A HIERARCHY OF TRADING DECISIONS

STEP 3.3c: TOUGH: URGENT TRADES

At last we come to the extreme application of trading skills: completing the trade at all costs because not getting the task accomplished is inadmissible. This type of trading requires that a trader grab liquidity, advertise trading interest to smoke out the other side, use the very best brokers, and remain assured that “at any cost” continues to be what the portfolio manager really wants. Here, more than ever, knowing who among your trusted brokers can be of help is critical. CONCLUSIONS

The hierarchy outlined above occurs before trading starts. The framework identifies those trades in the day’s workload that need immediate attention and/or require a high level of trading skills. Next the pick and shovel work starts, and the strategic plans will be modified as the individual traders interact with the market. Traders need to constantly test whether the trading reality matches the expectation and whether the chosen strategy continues to be the most sensible tactic. Above all, trading is an intensely competitive and adversarial activity. Not only do others want to win by WINTER 2006

Page 5

LY

is not well prescribed. A process is needed—gaining an acceptance or understanding to derive an outcome that is acceptable to both buyer and seller. The difference is that a process recognizes the uncertainty in the activity; what was the correct procedure at one moment may need to be abandoned for an alternate. As I said earlier, this structure is only one way to view the problem. It may represent a starting point for a trade desk to discuss how their procedures, and most importantly their adaptivity, could be improved. The decision tree is deceptively simple, but the key trader skills are awareness and adaptability. Excellence requires calling out the right skill at the right time while continually self-evaluating and adapting. It is this striving for excellence, using all available resources to economically implement portfolio managers’ ideas, that makes for best execution. Best execution is a process, not an outcome. Rather than being defined by the pieces in the puzzle box, best execution is the process by which the pieces are selected and the trading puzzle is assembled. To order reprints of this article, please contact Dewey Palmieri at [email protected] or 212-224-3675.

IE

trading against you; they also want to win by trading in anticipation of you. So managing the release of information and signals about trading intentions is as important as managing the process. Top trading resources are always a scarce quantity within an investment management shop. It is important that those best resources pay attention where it will make a difference: the tough trades. The first task is to identify those critical trades and get cracking on them. Another key point is that the different trading needs—order by order—will route to different trading destinations. This lies in stark contrast to simplistic concepts such as “Why don’t we do all our trades at a penny a share?” The answer to this silly question is that the brokers and trading venues capable of handling difficult trades won’t work for a penny a share. Pennies might buy you a rowboat to carry small packages, but that won’t do when you need a tugboat to move heavy materials. The critical distinction is between process and procedure. A procedure attempts to lay out in advance a series of steps to accomplish an end. Its virtue is its rigidity, accountability, and predictability for well-prescribed situations. Trading, like investment management in general,

N

5:58 PM

O

1/4/06

W

Jot_Wagner.qxp

EV

APPENDIX

R

Buy/Sell N-thousand Shares of XYZ stock

3.0 Unrestricted

FO

R

2.0 Restricted

3.3 Tough

Designated Broker

3.2 Easy

3.1 Liquidity Providing All venues, limit orders, unhurried brokers

3.2a Not Softable

3.2a Softable Soft Dollar Brokers

3.3a Large

3.3b Momentum Principals, DMA access

3.2b Crossable

3.2b Not Crossable

Crossing Venues

Auto-execution venues

3.3c Large, Urgent & Momentum Principals, DMA access

3.3a Crossable Crossing Venues

WINTER 2006

3.3b Not Crossable

Dealers, 3rd mkt brokers

THE JOURNAL OF TRADING

5

Jot_Wagner.qxp

1/4/06

5:58 PM

Page 6

Complete Decision Hierarchy Trade characteristics

Venues

Skills and systems

1.0 Order validation 2.0 Restricted

Designated broker

Accuracy

All venues

Patience

Soft dollar broker or step-out

Accounting, supervision

Crossing networks if possible

Versatile routing and control systems

DOT,SOES Algorithm, Program

Versatile routing and control systems

3.0 Unrestricted Liquidity providing

3.2.

Easy 3.2.a Softable

LY

3.2.b Non-softable Crossable Non-crossable 3.3 Tough 3.3.a Large Crossable Non-crossable

O

Crossing networks if possible

N

3.1.

Dealers, third market brokers

Well-nurtured relationship

Principals, DMA access

Attentiveness

Principals, DMA access

Aggressiveness

W

3.3.b Momentum

FO

R

R

EV

IE

3.3.c Urgent

Versatile routing and control systems

6

CREATING A HIERARCHY OF TRADING DECISIONS

WINTER 2006