I don’t think Sidney Crosby would ever say anything publically about topics as gauche as money or escrow, but his two training buddies pictured above occasionally complain that escrow is too high. It’s actually too low, but it’s understandable they feel the way they do.

Before the Pandemic and the CBA extension that quantified cap increases and escrow, we used to get news in early June on the official salary cap projections from the league. Next would come a period of speculation as the NHLPA voted on using the “escalator clause”, aka the inflator, which allowed them to goose the salary cap by as much at 5%, depending on how optimistic they were about the economic future of the NHL.

For many years they always voted yes to the full 5%. Numbers from before 2010 are harder to find, but when the inflator was set at 3.47% in 2015, it was said that was the second lowest since the CBA began in 2005. It not necessarily optimism that drove the players, it’s also older players nearing or at UFA age dominating the vote. It’s in their personal interest to vote for the inflator if they’re heading into free agency, and the players on ELCs don’t really care, their contracts are capped. The NHL was always too optimistic on forecasting future revenue, the players kept adding 5% to the bad guesses, escrow rose and the amount returned to players fell. Eventually the urge to line their own pockets met the reality that the contracts they were signing were illusory numbers they’d never see in their bank accounts.

Escrow in 7 points

Before we go any further, it’s time to re-learn what escrow is. This is how the NHL used to set the salary cap and the escrow pool that allowed them to reconcile their projections with ultimate reality:

  1. Sometime in June, the NHL sets a cap for each team based on their estimate of expected revenue growth over the current year’s actual figures, which they have in draft form by the summer. The NHL year is July 1 to June 30.
  2. The NHLPA votes to inflate that salary cap by an amount up to 5%. The bigger the cap, the bigger the free agent signings, and this vote is usually a few weeks before July 1.
  3. The final cap number is announced and then, before the new season starts, an escrow % is agreed upon between the NHL and the NHLPA.
  4. Escrow comes off of every NHL players’ salary cheques for the entire season, and then it sits in a bank account, idling away, usually for almost a full calendar year.
  5. After June 30, the NHL teams have to report their Hockey Related Revenue (HRR), as outlined in the NHL CBA to the league. The league assembles all those figures, confirms the various teams are doing it accurately and correctly and figures out the total HRR that was actually earned. This legitimately takes a lot of time. And the NHLPA has to sign off on all of this as well.
  6. The full HRR number is split 50-50, and if that 50% number is at or more than the salary paid out, everyone is thrilled, and the players get all of their escrow back. Not only that, the owners have to top the pool up with some extra money. If the 50% number is less than the revenue estimate, only some of the escrow goes to the players with the rest going to the owners.
  7. A full year after the hockey season ends, everyone has in their hands their fair share of the NHL’s actual earnings from that year.

This is what actually happened from 2010 to today:

In 2010, five years after this process had begun, the NHL’s revenue estimates were dead on, the players were voting every year for the full 5% escalator, and the result was that some of a very low escrow percentage was never returned. For 2010 through 2012, the amount of escrow not given back was entirely the result of the players themselves voting for the escalator.

2012-2013 is an odd year because the new CBA, post partial lockout, lowered the actual salary cap from $64.3 million to $60 million. The negotiation included an escalator of the full 5%, and the resulting very large escrow % was almost entirely used up. The projection was high, and the escalator made it worse.

You would expect a lesson to be learned there, but nope. For the next two years* the full 5% was applied, escrow stayed at or near 15% and the amount returned fell some more.

* Firm confirmation of the escalator number for those two years was very hard to find. Several later reports implied the full 5% had been used.

However, the real culprit in the high escrow and low amount returned was now the NHL’s revenue projections, not the players voting to up the salary cap, although that made a bad situation worse.

In 2015, the players finally voted for the second lowest escalator ever at 3.47% It didn’t really help much. Escrow was 17% and the return was low making the net retained by the owners the second highest since 2010. So lessons were learned? Nope.

One thing to remember is that the escrow % is set closer to the start of the season than the salary cap itself, so by the time that number is determined, and it can fluctuate through the season, the NHL has a clearer idea on real revenue for the season they’re in. It’s usually very realistic in that very little gets returned.

In 2016, the players got back on the 5% horse for a year, and did not prosper because of it. The change came in 2017, and the impetus was multi-fold. There had been concerns for a few years that the Canadian dollar was weak and driving down real revenues. The NHLPA head Donald Fehr was under pressure from a group of agents who thought the PA was being run poorly. And the NHL just really, really believed they could keep putting the cap up by two or three million a year, and the money would come from somewhere. Something had to give, and the slightly surprising result was that the players and the PA were suddenly the most fiscally responsible party involved and they began a period of voting in barely perceptible escalator amounts — 1.5%, 1.25% and then 0.5% in 2019.

The players had been forced to learn that there was no point to treating the NHL revenue projections with optimism. And this is where the real source of the ire lies, I believe. The escrow % stayed at around 12% and the amount returned stayed flat. They’d stopped topping up the cap and nothing got better! Their contracts were full of fake numbers they’d never get paid.

If the world had continued on past the winter of 2020 on the path it had been on, and the NHL had brought in new TV deals in a climate of rising ticket sales and happy fans, maybe this would be sorted out by now. Or maybe the NHL would still be guessing wrong. The world didn’t continue on unchanged, and things got massively worse on the escrow front. Escrow soared, nothing was returned, and the players are in debt to the escrow fund now with an amount that will take years to be paid off.

That’s not actually the NHL’s fault, though. But from a player point of view, it’s a little understandable if it feels like it is. Agents play into this and ironically tell the players the salary they “lose” to escrow is terrible. I say ironically because of course agents take a % too, to tell them the NHL’s cut is too big — amongst other things.

What a player sees is just another tax. It effectively is one, it comes off their salary, it gets paid into a fund, just like the income tax deducted, and long after the season is over, someone else decides how much they get back. Just like a tax. Income tax is deducted every paycheque because the government likes to get their money sooner, escrow is there to make sure the salary received wasn’t more than 50% of the real revenues.

This is where the Pandemic disaster to NHL salaries is arguably the league’s fault. The pain for the players would be less if the MOU had allowed for a split of revenues to be in a ratio other than 50-50. If the owners had taken less, the players would be less in debt. Understandably, the players like that idea. A lot of fans like it too, because yelling about billionaire owners is lucrative in the likes marketplace. However, if teams that were already cutting salaries, hours and positions for their non-player workforce suddenly had even less money, those cuts would be deeper. From that point of view, flowing more salary to players isn’t really a power to the worker stance it seems to be.

This brings us to that MOU with it’s mandated escrow numbers that were extremely optimistic in terms of how fast the NHL could rebound with ticket sales and other revenue. So far, that optimism seems to be paying off. There’s been some disruption to this season, but not a lot, and the TV ratings for some of the new ESPN broadcasts are very favourable. The Rangers going deep in the playoffs should make an impact on HRR that will help drain the escrow debt a little faster. But the low escrow numbers now mean players of the future will still be paying off this debt and the cap will stay flat for at least two more seasons after next.

I’d worry this successfully optimistic revenue forecasting would lead the NHL back into bad habits, but the MOU has a corrective. Once the salary cap can be calculated by HRR again — so whenever that debt is paid off — the prior three years are used to form the projection, which should slow the rush to assume growth in advance. Until then, the cap will go up by only one or two million per year.

In other words — this problem is actually fixed in the only way it can be fixed — by having the revenue projection closer to reality like it was in 2010. As long as the players stay away from the temptation of the escalator (stairs are better cardio) they’ll be fine. And likely they’ll stop complaining and move onto some other self-centred agenda, just like the rest of us are wont to do.