Beruflich Dokumente
Kultur Dokumente
Limit Orders
Compare market orders with limit orders, including the price and execution uncertainty of each
If you absolutely have to get a trade done right away, you go with a market order.
Let's say you want 10,000 shares of ABC at a price of $20/share.
You might not get exactly the price that you would prefer (ie. you will likely pay above
$20/share).
However, your order is almost certain to be filled (ie. you will likely get all 10,000 shares).
But, what if I care more about getting the price I want than I do about getting my whole order filled?
Calculate and interpret the effective spread of a market order and contrast it to the quoted bid–ask
spread as a measure of trading cost
This is simply the difference between the bid and ask prices and is also know as the inside bid-
ask spread, inside spread, market bid-ask spread, and market spread.
What you may not have know about before is that the halfway point between the bid and ask
prices is known as the "midquote".
So then, what is the "effective spread"?
First, you need to realize that the bid/ask spread is a flawed measure because sales often take
place at prices above the bid price or below the ask price.
To get the effective spread, simply take the difference between the execution price and the
midquote from just before the trade occurred and multiply by 2.
Multiply by 2?
Yes, which means if you see two possible multiple choice answers, the one that is exactly double
the other one is probably correct.
What is the effective spread when an order is carried out at more than one execution price?
It sounds like you're saying that the effective spread is a better measure than the quoted spread.
If the effective spread is lower than the quoted spread, that's what's happening.
Actually, yes.
1. Electronic crossing networks: Investors place orders, which are then matched (or crossed) based
on averaged bid/ask prices at set intervals.
2. Auction markets: traders compete against each other to fulfill orders.
3. Automated auctions: Computerized auction markets (also known as electronic limit order
markets).
This is just a name for market that have characteristics of more than one market type.
Brokers are hired by the investor and work FOR the investor.
In exchange for a commission, brokers provides the following services:
Dealers act AGAINST the investor by under-bidding when the investor wants to sell and over-
asking when the investor want to buy.
Dealers are not agents, rather they are in business for themselves and receive compensation in
the form of trading gains.
However, Dealers are not entirely bad because they do provide liquidity to the market.
Also, you as a trader are more likely to want to buy or sell when you have an information
advantage over the Dealer (or think that you do). This means that the Dealer has "Adverse
Selection Risk".
Not necessarily, there are also "broker/dealers", which is when a broker decides that it is more
likely to be successful if it participates in the transaction.
In such a case, the broker effectively becomes your minority partner and captures some of the
upside in your deal (ie. they put up some of their own money and get more than just a
commission).
Explain the criteria of market quality and evaluate the quality of a market when given a description of its
characteristics
1) Liqidity
2) Transparency
Investors need to have confidence that, when they make a trade, the counterparty will uphold
its obligations (ie. to pay or deliver)
This is why, for example, clearing houses act as guarantors of performance.
Okay, that all seems pretty straightforward, how is this likely to show up on the exam?
I see this as much more likely to show up in the afternoon session as part of a multiple choice
question set.
Take a look at Example 3, where they give you an observation about what had happened to a
market in terms of its quoted spreads or quoted depths, or some such characteristic, and you
need to determine whether you think that this improves or deteriorates market quality.
Also, while there are three characteristics mentioned in this LOS, the text devotes an entire page
to discussing liquidity, which is not particularly lengthy in the big picture, but transparency and
execution/assurity of completion are forced to share a small paragraph on the next page, so
chances are that you will need to know a) what characteristics make a market more or less
liquid, and b) liquidity is ALWAYS GOOD.
Explain the components of execution costs, including explicit and implicit costs, and evaluate a trade in
terms of these costs
What exactly are the components of execution costs? Aren't these just transaction fees?
These are easily quantifiable because they have a price tag (commissions, fees, taxes, etc.).
Wait. You just said that execution costs weren't about transaction fees.
Which means?
Which means that you don't actually get a bill for them, but they do exist.
You just have to work through some annoying calculations in order to find them.
These are discussed in 30g.
Annoying calculations? I'm going to not learn those and hope this doesn't show up on the exam.
Not a wise move because these are pretty much the most important part of this Reading and
there's a good chance that it will show up on the exam.
What would cause this to happen? Why can't people just buy at the ask or sell at the bid?
And the other implicit execution costs are... Wait, what did you just say?
Moving on.
This is the gain that you miss out on (or loss you avoid) if your order does not get completely
filled.
This typically happens with limit orders, which have execution uncertainty.
This is measured using the difference between some closing price and the benchmark price,
which is the price at which the manager decided he wanted to buy it (or sell, but on the exam it
will almost certainly be a buying scenario).
4) Delay/Slippage costs
These costs are a lot like missed opportunity trading costs in the sense that they arise when an
order cannot be completed immediately.
The difference is that missed opportunity trading costs arise when an investor is holding out for
a specific price. It's not that she is prevented from trading, it's that she won't trade at a price
beyond her limit.
In the case of delay/slippage costs, the market (rather than the investor) is the limiting factor
because it cannot provide sufficient liquidity to process the order.
This typically arises when an order to particularly large and must be broken down into smaller
lots in order to be filled.
Because this can take a while, information is leaking out to the market and the price starts to
move away from where it was when the investor made the initial decision to trade.
What are the prices that we need to know to calculate implementation shortfall?
Benchmark price (BP): The price that the manager sees and decides to buy/sell at (assume that
it's the closing price on the first day of a case/question set).
Decision price (DP): This is the closing price on the day before any part of the order gets filled, so
it will be different for parts of the order that are filled on different days.
Wait, the "benchmark price" is the price that the manager sees and decides to buy/sell at, but the
"decision price" is something else?
Whatever. If they try to pull crap like that on exam day, I'm going to start a riot. Anyway, what are the
other prices that we need to know to calculate implementation shortfall?
Execution price (EP): The actual transaction price (for the portion of the order that actually gets
filled).
Cancellation price (CP): The closing price on the day that the order is cancelled and the
remaining portion is unfilled.
The formulas below provide each component of implementation shortfall as a % of order, but
you may also be asked for these expressed in currency terms, which you can do by multiplying
the %s by the amount of the original order.
These formulas are all different. Can you help me understand them a bit better?
Start with the denominator, which is always the benchmark price.
Next, note that delay costs and realized profit/loss are multiplied by the % of the order that is
actually filled - not the entire order size and not the % that goes unfilled.
Missed opportunity costs are multiplied by only the % of the order that does NOT get filled,
which makes sense.
Explicit costs, by contrast, are applied to the entire order - so $20 trading fee on an order for
1,000 shares is $0.02/share - even if the order does not get completely filled.
Those are all logical, but the prices used in the numerator are trickier:
Delay/Slippage costs start with DP - BP (think "Delay means use DP")
Missed opportunity costs start with CP - BP, which makes sense because the you pretty
much admit you've lost the opportunity when you cancel the rest of your order
(although remember that there is no set rule on what actually is the "cancellation
price").
Realized profit/loss starts with EC - DP, which actually makes sense - just remember that
the DP is used in the numerator, but it is still the BP that is used in the denominator.
I'm not really feeling like putting in much of an effort today, so I'm basically just going to give
you the example from Section 3.1 with tiny changes.
Order placed for 1,000 shares, but only 60% of the order was filled.
Yes.
Unlike pre-trade costs (see 30i), Implementation Shortfall (and especially Market-Adjusted
Implementation Shortfall) can be negative, which means that these costs can actually end up
being a net benefit to you.
Think about the big picture - the main point of this Reading is to teach you that implicit trading
costs are typically much greater than explicit trading costs. See the iceberg picture in Exhibit 7.
Also, always remember that there is a considerable amount of ambiguity about these
calculations because it is not exactly clear which price should be used as the benchmark price,
etc. So they will probably tell you exactly which price to use as the benchmark price.
Contrast volume weighted average price (VWAP) and implementation shortfall as measures of
transaction costs
Volume-weighted average price, or VWAP, is a measure that is used when you can't complete a
transaction at a single price.
For example, you buy 100 shares at $20.00, 250 at $20.01, 500 at $20.02, etc.
VWAP gives you the "true" price of the transaction by weighting the portions sold at various
prices.
If the primary advantages are "easy to calculate" and "easy to understand", then it must really suck.
Excellent observation. It does.
Specifically, it fails to account for market impact (which we discussed in30f).
Also, it fails to account for all of the costs associated with trading, such as delays or unfilled
portions of orders.
Finally, traders can game it.
The primary advantage is that traders can't game it (well, I'm not so sure about this, bu that's
what the Reading says, so assume that it's true).
Also, it breaks trading cost down into its component parts and allows investors to analyze and
minimize trading costs.
Finally, it allows investors to see the trade-off between quick execution (realized profit/loss) and
market impact.
Because the calculations are annoying and the definitions are ambiguous.
Also, they are the very people who stand the most to lose if everyone moves from VWAP to IS.
Explain the use of econometric methods in pretrade analysis to estimate implicit transaction costs
Stop right there. Pretend that it is only a couple of days until the exam and I don't care about your
massive table. Tell me what I absolutely HAVE to know. Then you can show me your stupid table.
Fair enough, I will "pretend" that the exam is just a couple days away.
However, it will require me to provide you with another (smaller) table.
Thanks, but how will I know what is a high or low trade size?
What combination of size and urgency would make me want to use VWAP as a tactic?
Explain the motivation for algorithmic trading and discuss the basic classes of algorithmic trading
strategies
Automated, electronic models that trade according to established quantitative rules within user-
defined constraints.
Probably somewhere in the range of 30% - 40% of trading is automated (ie. conducted by
algorithms).
Ideally, algorithms should be able to identify market trends and seize upon even the smallest
and briefest mispricing.
The primary advantage is that computers won't commit human error and can minimize trading
costs.
Also, algorithms can break conceal large block trades by breaking them up into small chunks and
reducing market impact (Meat Grinder Effect).
This strategy can yield large savings compared to making a single block-trade.
Logical Participation Strategies are designed to get an order filled while minimizing trading costs.
Simple Logical Participation Strategies involve participating in the market, but not standing out. There
are three sub-styles:
VWAP strategy: Match or beat VWAP for the day (Most popular Algorithmic Trading Strategy)
TWAP strategy: Trade at a constant volume and match or beat TWAP (good for thinly-traded
securities)
Percent of Volume strategy: Trade a security at a percent of overall volume until your trade is
complete
Implementation Shortfall Strategies also seek to fill an order, but minimize trading costs based on
Implementation Shortfall calculations (as opposed to VWAP or TWAP)
Opportunistic Strategies are not designed to fill orders, but rather to seize upon opportunities
such as temporarily liquidity or mispricing.
Passive trading combined with seizing opportunities for high liquidity
Reserve (Hidden) orders and crossing are used in an attempt to achieve negative trading costs
Specialized Strategies
Include, but are not limited to, Passive order strategies, Hunter strategies, and a few other
strategies that have approximately 0% chance of being tested on the exam.
Discuss the factors that typically determine the selection of a specific algorithmic trading strategy,
including:
order size,
Translating CFA-speak
A big theme in this reading is trying to minimize the effect that your trade has on the market. This effect
is called Market Impact and was first discussed in LOS30b.
Basically, there are quoted bid and ask prices, but the order sizes behind these prices might be very
small compared to your order and you'll have to fill it at successively less attractive prices. In the
reading, this is referred to "an immediate demand for a large amount of liquidity."
The idea behind Algorithmic Trading strategies is to move a chunk of your transaction at the attractive
price (bid or ask depending on whether you're selling or buying) and then waiting until someone else fills
the void that has been created by your counterparty getting out of the market.
Traders must employ the Algorithmic Trading strategy that is best suited to current market conditions.
In order to accomplish this, they will compare potential trades according to factors such as:
Based on these factors, traders can identify an ideal Algorithmic Trading strategy. For example:
"Best Execution" is like "Prudence" (Standard 3a), in that it is your duty to pursue it, and you
know when it hasn't been done, but it is virtually impossible to define
Best execution is impossible to define objectively because some trades may have high
transaction costs, but be the right trade in the overall portfolio context
Best execution can't be known in advance of a trade because each trade takes place under
unique circumstances
Best execution can be measured after the fact, but trades may take place in extreme market
conditions
Best execution is an ongoing process that is continually refined, so we have to examine it in a
long-term context
Evaluate a firm’s investment and trading procedures, including processes, disclosures, and record
keeping, with respect to best execution;
Yes.
Yes.
Yes.
Clearly, this is a radical departure from every other piece of advice in the curriculum.
30p Fiduciary duties: Trading execution
In theory.