2
$\begingroup$

When an institutional trader wants to enter or exit a large position, the broker has to process the trade in smaller chunks to keep market impact at a minimum. So the broker would say that they can take the position over a certain amount of time and guarantee the Volume Weighted Average Price for the shares bought or sold over that time.

That is why intraday Volume Weighted Average Price oftentimes acts as a fairly robust support or resistance level, since a large volume of trades happens as the price approaches it.

I wonder, are there any other calculable price levels or parameters like the Volume Weighted Average Price which are used by Brokers to process large institutional orders and therefore might have consistent relevance?

$\endgroup$
3
$\begingroup$

I think you are misunderstanding. Volume Weighted Average Price (VWAP) is both an execution tactic and a benchmark. VWAP as a benchmark simply means the [gross notional traded during the day] / [day volume]. Special trades (derriv tied, as-of, etc) are excluded. VWAP is basically the weighted average trade price of the day that you could have been able to achieve yourself.

Institutional clients will often send their orders to brokers' algos with the instruction to target the VWAP. That just means that the client wants to sort of execute in-line and doesn't have much of a view as far as execution strategy. VWAP as a tactic means that the broker's algo should trade trying to execute percentage-wise the same amount as will trade in the market overall. For example, if you have an order for 1,000,000 shares of XYZ, and on-average 10% of the volume in XYZ trades between 10:30am and 11:30 am, then the algo will execute 100,000 shares in that time period.

Brokers offer many different algos. Some are simple, like TWAP. Time Weighted Average rice. That just means take the number of shares, the time window to execute, and then split it up evenly.

There are all sorts of other algos that do things like trying to minimize impact or take advantage of block liquidity when it becomes available.

On the research side you need to calculate a return series based on something. You could go from open to open, close to close, close to next-day open, etc. Data mining is endless. Another statistic that people use is the VWAP, since its just the average price that traded that day.

There's really no magic to it. It's not that a large volume of trades happen as the price approaches it - it's that by definition the VWAP is the volume-weighted average.

Does that help?

$\endgroup$
  • $\begingroup$ Yeah, that makes sense. However, I often see e.g. on SPY that the price touches VWAP and reflects off of it throughout a day, implying that activity increases as it approaches that value (which steadily changes over time). Is that just a fluke? $\endgroup$ – Kagaratsch Aug 30 at 2:25
  • 1
    $\begingroup$ Yeah, the vwap is only known after the final trade. Each trade affects the calculation of VWAP. And SPY interacts with the price of the futures, which is somewhat disconnected from the ETF... $\endgroup$ – JoshK Aug 30 at 2:45
  • $\begingroup$ @JoshK That is very interesting! Could you tell me more about SPYs interaction with futures? $\endgroup$ – Kagaratsch Aug 30 at 11:03
  • $\begingroup$ Sure, I'd be happy to. But I think it might be better to set up a new question, since it will be more text than can fit into the comments. $\endgroup$ – JoshK Aug 30 at 14:49
  • $\begingroup$ I created a new question about it here: quant.stackexchange.com/questions/47423/… $\endgroup$ – Kagaratsch Aug 31 at 2:30
2
$\begingroup$

Question is moreover of Portfolio Trading Strategies

For a Portfolio of Stocks to be traded with both buys and sells, one must consider the trade basket as a whole. For instance, an imbalance between buys and sells might cause an intended net market exposure. The correlation between the stocks is another important issue. For buys and sells that are highly correlated in terms of stocks returns, one would like to synchronize the trades, because doing so would reduce systematic exposure. However, if these trades have different market impacts, one would like to execute them at different speeds to minimize the transaction cost. It is therefore necessary to find the balance between two.

The trading horizon --- the length of time we allocate to implement the trades --- is the another important factor. For trades that are easy to implement based on liquidity, the trading horizon should be short. For difficult trades, the trading horizons can be longer. For a given set of trades, it is better to optimize the trading horizon as well as the actual trade implementation.

All Institutions have models in place to find out the

Optimal Solution with Fixed Trading Horizon is an example

Mathematical technique to solve above type of optimization problem is the calculus of variation.

Please note when trading a Portfolio of stocks, one often has to maintain the balance between orders so that portfolio meets set of constraints.

All orders are placed at Limit order only.


$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.