I commented on the original question
Don't do it. No results means no abstract.
While this received many upticks, I have also been told
That sounds incorrect. Any references or related experience?
and that statement also has some upticks. While this is not an answer to "how to write an abstract" it attempts to clarify my comment (but is too long to be another comment). Hopefully it is helpful.
I think there are so many things wrong with writing an abstract without results that it is difficult to explain my thinking. The apparent reason for wanting to write an abstract without any results is
No abstract means missed opportunity; I cannot afford that. I have to present results I'm sure I will have by the time of the conference.
Which has recieved essentially the same number of upticks as my comment not to do it. I disagree with "I cannot afford that". I have never seen or heard of someone being denied tenure or a job because they didn't present at a conference one year. Hiring decisions are never so close that a single conference presentation (no matter how prestigious) sways the decision. I would argue that there are very few upsides to submitting an abstract without results and potentially some downsides.
Submitting an abstract without any results will not get you a place at a highly prestigious conference or a keynote address. It will get you a place at a conference that essentially accepts all abstracts, but not much more than that. In fields that I am familiar with conferences happen at least every 6 months and more often every 3 months. This means that by not submitting now you are merely delaying your presentation by 3-6 months. Therefore the cost of not submitting is a 6 month delay and a slightly different conference that is potentially slightly more prestigious (e.g., with results you might be able to get a talk instead of a poster).
In slow moving fields 6 months is essentially meaningless. In fast moving fields, 6 months is a long time, but in the fast moving fields I am aware of you don't present results until they are about to be published. This means you don't want to submit an abstract of results you don't yet have. Therefore I see very little cost of waiting for the next conference.
So what are the benefits of waiting. Again they are not great. The abstract will actually represent what you are going to talk about. You will likely get put in the correct session. There is a higher chance of getting a talk. If everything goes tits up, you will not have to withdraw. While most people will not remember, some of your close colleagues will and this could hurt future references. Withdrawing also screws over the conference organizers and they will not forget.
There is also the issue of how long do you need to get results. If the abstract was due the day of the conference, presumably you would want to have results before submitting the abstract. What about a week? A month? 6 months? Where is the line?
Finally there is the issue of integrity. While one can write the abstract to make no promises and only state the current truth, this is in fact difficult. If you do this frequently enough you will likely eventually make a statement that is a lie.
In an attempt to answer the question, what about:
We don't have any results yet as it is still N months before the conference. By the time the conference rolls around we are sure we will have something interesting. If not, we will present some old data or just not show up.