How to set a sell limit order that will actually execute?

Question

I'm having trouble getting a good answer to this with Google. I'm also new to this. Haven't actually tried it yet.

I have several hundred shares of a security that I'm hoping to sell.

Let's say bids are around $34, asks are around $41, last trade price is around $38. Based on my Googling, this is wider than the examples I've seen. Volume is around 19k.

I'm not in a hurry. As I've read, I want the price more than to move out of the position quickly, thus the limit order.

What I can't seem to find a clear answer on is where I put my sell limit order price to give it a chance of executing.

Would I set an ask limit order near the ask price? (If it's $41.00, somewhere between, say $40.50 and $41.50?)

Do I need to sit closer to last trade price?

All the way down to bid?

Answer 1

There is no perfect answer to this question, and you will always have a tradeoff between likelihood of executing the order and the price you wish to achieve.

The only way to be certain to execute the order is to sell at the bid (or more accurately at the highest bid that immediately accommodates the full volume you are trying to sell), and the market's price can move away from you as easily as towards you.

Depending on your broker, there may be liquidity they can access that isn't immediately obvious to you (so-called "dark pools"), so it may be worth trying a high initial bid, just under the ask... Your broker may also have algorithms that purport to execute inside the spread only: IB has tried a couple ways of doing this, but my limited experience with these algorithms hasn't been stellar (though I haven't tried their newest version).

My approach for wide spreads if I need to sell but am in no particular rush is to set my limit just under the current ask, wait 5 minutes and look. Often a new ask will have appeared just under me. Rinse and repeat until you converge on the minimum price the other potential seller is willing to settle for... then wait. Towards the end of the day, start lowering towards the bid if you really need to sell. This approach doesn't guarantee anything except that if you execute, it will be at a "reasonable" price (in the sense that you will have aided price discovery and added liquidity to the market... anything inside the spread is objectively reasonable in that sense). Setting your limit outside the spread, especially when the spread is wide, almost guarantees that you won't execute in my experience. The exception is when your security usually trades only in very large chunks, but this happens usually only with some weird ETNs, and even then you're gambling that the next large liquidity taking order will be the opposite direction you are trying for.

The way that vanilla long-only equity fund managers usually do this is to request "VWAP" from their brokers. VWAP stands for Volume Weighted Average Price, and is essentially the price you'd get over a period of time if you took a proportion of each trade. Their brokers will then do their best efforts achieve that price, essentially by trying to be part of every trade during the time period, in the correct proportion. They have to forecast things like tick-by-tick volatility and depth. It's not something that you or I could do at home. They usually do better than we as individual investors could do, but the difference is not as big as you'd think, and in less liquid stocks sometimes the best achievable is over a percent worse than VWAP. I don't think they're ever trading stocks quite this illiquid, though...

It's worth keeping in mind that this whole game is just that... price discovery is exactly what the HFT people work so hard on, and it's how they make their money. Their raison d'etre is narrowing spreads, and if they have not succeeded in narrowing the spread in the security you are trying to trade, it's almost certainly not worth your effort to try to beat them here.

Answer 2

$34 x $41 is an absurd spread. Is this an actual real time quote or is this 'market is closed' quote after traders have pulled their bids?

A limit order is an order to buy or sell a security at a given price or better. If the price selected is better than the market, it will not be filled until price reaches the limit. That may or may not happen.

The current B/A is where the market is right now ($34 x $41). If you want a 100% of a fill right now, sell at the $34 bid. If you set your limit at $35, the chance of executing is better than if you set it at $36, and so on at $37, $38, etc. If you want a better price, you'll need a counter party willing to pay that price. When/if this will occur is unknown.

Since you have indicated that you are willing to hold the security, just set a target sale price that is acceptable. Reevaluate if there is adverse news and change the limit price if need be.

For example, if $40 is acceptable to you and if your broker offers more sophisticated algorithms, set up a self adjusting PEG order that lowers your sell price down to your limit price. IOW, the best ask price is current $41. Set your sell price at $40.99 and every time someone comes in at a better offer price, your PEG order will lower your sell price until it reaches $40, at which time, it will not lower price any more.

Answer 3

One thing I have done that you could try is to set the limit at the ask, leave it there for a while, cancel the order, bump up your price down little on a new order, and so on until you get a fill. If the stock is trading a lot at the ask you could eventually wind up at the front of the line and get a fill. Although with that kind of spread I suspect nothing is trading at all. In such a case you might not get a fill at anything other than the bid, at least not immediately.

One thing to note is that whatever you set your limit order for should become the new ask with a depth at whatever number of shares you want to buy at, because you are basically asking less than all the other sellers at the time. So it's possible that whatever you ask may eventually attract a buyer's attention, assuming another better offer isn't placed (by a person, trading algorithm, etc.) in the mean time.