In theory, racehorse breeders are all striving for a similar primary goal: foals that will go on to successful race careers. In practice, it's not so simple.
Here's one of racing's open secrets: there are stallions that are primarily "commercial," and others that are "breed-to-race."
We'll leave aside - for today - the discussion of commercial breeding, and focus on what paints a stud as a breed-to-race sire.
Stud fee is one component. Investing six figures on a stud fee is a commercial sales strategy; breeders who race their foals must consider potential racetrack profitability. The ideal breed-to-race stallion will stand for no more than $15,000 - and preferably $7,500 or less.
These sires must get consistently good offspring. Let's say they have at least a 50% winners-to-foals ratio, and probably 70% or better winners-to-starters. Average earnings per starter should top $50,000 (generally a break-even amount).
Notice that I said "consistently good" offspring. One knock against breed-to-race sires is that they rarely come up with a "big horse" that catches the eye of a commercial market (when they do, their raised stud fee removes them from the breed-to-race category). So I'll add that these stallions should produce at least 5% stakes winners, but I won't include a threshold for graded winners.
Durability is another factor. If the resulting foal needs to earn his way (and provide a profit for the owner) during his race career, the odds are improved if it's a long career with many races. The national average of starts-per-starter is 15 and change... I'll peg the breed-to-race figure at 17 and up.
A stallion that fits the bill for owners who run homebreds must improve his mares. We all know the rule: AEI must be fairly high, and must exceed CI. Since AEI and CI are indexed at 1.0, I'll say the difference should be 0.2 or better (i.e., the stallion significantly improves his mares - he doesn't benefit from "mating up" to higher-class broodmares).
Tomorrow, I'll post a list of my "top picks" for breed-to-race sires. I'll give you a day to think about it and see if you agree with my selections or -- I suspect! -- have a few of your own to share.
For today, I'd like to keep the comments focused on the factors that make for a breed-to-race sire instead of naming specific stallions. What can you add to what's above?