The End of an Era – Say Goodbye to All-Or-None Orders

A few days ago, the CBOE announced that it would no longer accept orders stamped with the “all-or-none” label beginning in September (please refer to CBOE Regulatory Circular # RG12-097, dated: July 16, 2012). All-Or-None orders are often refered to as AON orders.


When I first saw the news, I thought I was imagining things. All-or-none orders have been around for the longest time, and they had become part of the standard option vernacular. I began to imagine the backlash that might come from the trading community. Traders would be upset that they could not guarantee a fill on an exact number of contracts, brokers would be upset that there would be one less feature on their platform, and educators would be upset that there would be one less piece of minutia to teach!

And then I though…Is doing away with all-or-none orders really such a bad thing?

What Is an All-Or-None Order?

Before we begin, let’s quickly define an all-or-none order. As its name implies, it is a limit order in which a trader requests to be filled either on every contract or on no part of the order. In other words, the one thing the trader does not want is a partial fill.

What Can Be So Bad About That?

To explain the downside to all-or-none orders, let’s first understand how a normal market works. Suppose that the market (the bid-ask spread) for a particular option is:

2.60 – 3.00

In other words, an interested seller understands that the most someone is willing to pay for his option is 2.60. An interested buyer knows that the cheapest someone is willing to sell him some options is 3.00. Said another way, 2.60 is the best advertised bid price, while 3.00 is the best advertised ask price.

Now suppose that Trader Bob wants to pay 2.80 for 100 contracts. Said another way, Trader Bob is 2.80 bid for 100 options. If Trader Bob submits a limit order to buy 100 options for 2.80 and he is not immediately filled, then because 2.80 is a better bid than 2.60, which was the previous bid, it follows that the 2.80 bid will replace the 2.60 bid as the best available bid in the marketplace.

This makes the new market for our option:

2.80 – 3.00

That 2.80 bid is Trader Bob’s bid. Since it is better than any other available bid, it is reflected in the market, and to Trader Bob’s delight, it is advertised, thereby giving the order a greater chance of enticing a seller to fill the order.

But what if Trader Bob decided to send a limit order to pay 2.80 for 100 contracts and labeled the order “all-or-none”? Well, this is where things can take a turn for the worse. Let’s explain:

The Problem with All-Or-None Orders

Let’s begin once again with the original option market, which was:

2.60 – 3.00

Then, suppose Trader Bob decided to buy 100 contracts for 2.80 all-or-none.

Then, even though Bob’s 2.80 bid is superior to the current 2.60 bid, it is NOT reflected due to the all-or-none contingency. This is because a market that reads 2.80 – 3.00 implies that a trader can sell as little as one contract at 2.80. With an all-or-none contingency, this would not be possible.

Now that Bob’s AON order is not reflected, and hence, not advertised, to make matters worse, think about what would happen if an order came in to sell 75 contracts at market. Do you want to guess where they would trade? You guessed it… 2.60.

Suppose a second order came in to sell 75 contracts at market. You guessed it! They could trade at 2.60 as well. In fact, dozens of orders to sell 75 or 50 or 10 contracts at market could enter the marketplace and they would all likely be filled at prices worse than what Bob was willing to pay!

That leaves us with two big losers. The first one would be the option sellers who sold 80 options at 2.60 when they could have sold them at 2.80. The second is Trader Bob who did not obtain a fill on a single contract even though he was willing to pay significantly more than where the trades ended up being filled.

Some Concluding Remarks

There was a time when all-or-none orders served a purpose. Decades ago, some brokers charged a separate ticket charge for each partial fill of a trade as well as a per-contract fee. As such, a trader could fall prey to extremely high commission structures (Bob buying 10 contracts at $2.80, then 25 contracts fifteen minutes later at $2.80, etc). The all-or-none order served as a way to work around those excessive commissions. Nowadays, the standard for firms with a ticket charge is one ticket charge per-day, regardless of the number of partial fills it takes to complete an order. Of course, per contract fees still apply. Nevertheless, since the benefit provided by the all-or-none order no longer exists, there is no good reason a trader would want to be exposed to the unwelcome side effects discussed in this article.

Alex Mendoza