In 2016, we are well past the time when cord-cutting and streaming entered the mainstream. Why, then, is streaming that replicates TV still such a tough nut to crack?

Why can’t we simply head over to YouTube or Facebook Live—or something equally user-friendly—and watch SNL or The Daily Show or Atlanta streaming live with ads, just as on a regular broadcast? Obviously this is a complicated field. Here, for example, is The Economist’s recent summary of the state of play. But let me offer my two cents.

The technology certainly shouldn’t be a problem. Live streams of the presidential debates from the full alphabet of cable news networks were readily available on YouTube this year. Sling TV also proves that subscription-supported streaming TV can exist, at least in the short term, but the price and product are not at all comparable to either clicking on a YouTube link or subscribing to an on-demand service like Netflix.

Two big issues, it seems, are the difficulty in replicating per-subscriber fees to the network in the online setting, and the weird obsession with proprietary technology among the content providers rather than piggy-backing on established methods.

There are three possibilities:

1. Money is being left on the table

If live TV streaming would be profitable but isn’t being done, the blame would seem to fall on the weird market structure of television, with production companies, national networks, affiliates, and cable and satellite providers all in the mix. The complex nature of fees, bundling, and multi-party contracts make the problem difficult to analyze and solve.

And yet if there is enough cash going missing, you’d hope that standard Coaseian logic would find a way to realize the gains. Is the cost of navigating the various vested interests enough to outweigh the potential gains?

2. Money is not being left on the table

If ad-supported live TV streaming would not be profitable, then I’m not sure we really understand how to think about the ad-supported internet in general. Ad targeting and segmentation are, in the conventional wisdom, supposed to be easier and more effective on the internet than in “old media”, and so if no premium is being commanded then something seems to have gone quite wrong.

This raises uncomfortable questions about consumer preferences in streaming versus cable situations. Is the rate at which consumers are willing to substitute “watching ads” for “paying money” out of whack with the relative size of ad rates to subscription fees? Is the “online=free” psychology of streaming insurmountable?

3. We don’t know if money is being left on the table

Maybe the problem is inertia and uncertainty. I’m open to the idea that we might just not really know what would happen here. The NFL’s deal to stream live games on Twitter is one sign that we might start to see more experiments, but given the NFL’s general malaise this year it might be a tough one to disentangle.

In this bin we can also put device and app headaches. This is one place where Sling TV still lags cable—frictions in the consumption experience are simply higher when streaming through any number and variety of devices, through a stand-alone app rather than in    Cable TV, for all its warts, is a relatively friction-free experience after the initial setup, and streaming in general is still not good at delivering comparable user-friendliness and uptime for the average user. The attempts to replicate (and undercut) cable online are not truly substitutes.

In sum, the progression seems to have gone like this: cable TV gets expensive as streaming gets viable; on-demand ad-free streaming becomes the dominant model online as cable protects its old turf; it’s hard to sell an online cable replacement even if it’s cheaper because by now everyone’s paying for Netflix and Amazon Prime.

If ads alone really could support the business model, then it’s hard to see where the sticking point is. In that sense, I think this whole messy industry might reveal a lot about the state of online advertising in 2016.