Canada's public pension funds are changing the deal-making landscape;

They own assets all over the world, including property in Manhattan, utilities in Chile, international airports and the high-speed railway connecting London to the Channel tunnel. They have taken part in six of the top 100 leveraged buy-outs in history. They have won the attention both of Wall Street firms, which consider them rivals, and institutional investors, which aspire to be like them.

These giants are Canada's largest public pension-fund groups. They manage around C640 billion dollar(643 billion dollar) between them. Of the 40 largest public pension funds in the world, four are Canadian, according to Preqin, a research firm (see table). America aside, no other country has more on the list. But size is not what marks them out. Their approach to investment is intriguing.

Unlike those in charge of public pension funds elsewhere, the Canadians prefer to run their portfolios internally and invest directly. They put more of their money into buy-outs, infrastructure and property, believing that these produce higher returns than publicly traded stocks and bonds. They are in some ways like depoliticised sovereign-wealth funds. Jim Leech, the boss of Ontario Teachers' Pension Plan, calls them a “new brand of financial institution”. And as public pensions around the world cope with painfully low yields on their assets, many see them as a template. Michael Bloomberg, New York City's mayor, is among the model's admirers.

Ontario Teachers pioneered this style of investing in the 1990s when it brought more of its investments in-house. Other large funds soon followed suit, building up teams to handle deals on their own. The Ontario Municipal Employees Retirement System (OMERS) wants to have 90% of its assets managed internally by the end of 2012,leaving some room for allocations to external managers in specific areas.

The funds will do smaller deals alone but often act as “co-sponsors” with leading private-equity firms on bigger transactions. This allows them to have more control over the investment and to save on fees. CPP Investment Board and PSP Investments, for example, worked with Apax, a buy-out firm, to acquire Kinetic Concepts, an American medical-devices company, for 6.1 billion dollar. It was one of the biggest buy-out deals in 2011.

The main draw of the Canadian model is cost savings. Running operations directly helps plug “the incredible leakage that goes out through fees” to pricey external managers, says Gordon Fyfe, the boss of PSP Investments, a large fund. In private equity, for example, many managers charge a fee equal to 2% of assets and 20% of profits. Hiring staff and building up internal capabilities costs far less. Keith Ambachtsheer of KPA Advisory Services, a pensions consultancy, says running assets internally costs a tenth of what it would if they were outsourced.

There are other advantages. Public pension funds have a longer investment horizon than private-equity firms, so the Canadian behemoths can choose to sell when the time is right, mitigating some of the risks of investing in illiquid assets. In December, for example, Ontario Teachers announced it would sell its majority stake in Maple Leaf Sports and Entertainment, a large Canadian sports business, for around C1.3 billion dollar. They had been invested in Maple Leaf since 1994, far longer than a private-equity firm's typical five-year horizon, and are expected to get a return of five times their money.

Because they are saving so much on fees and only need to meet the liabilities of scheme-members' pensions, moreover, the Canadian funds feel less pressure to chase the high returns that leading buy-out firms do. They can pursue investments with less risk and leverage. “Because returns don't have to be as good, they can bid more for companies,” says one buy-out boss. It sounds like a loser's lament.

So far the funds' strategy has paid off. Over the past ten years Ontario Teachers has had the highest total returns of the biggest 330 public and private pension funds in the world. Some of this outperformance stems from the relative strength of Canadian stockmarkets and property, to which Canadian pension funds have higher allocations than others. But not all of it. In 2010 OMERS returned 25 dollar on every dollar spent on internal management, for instance, compared to only 10 dollar for every dollar spent on external managers' fees.

Those seeking to understand how Canadians have pulled it off are given two answers: governance and pay. There is little political interference in the funds' operations. They attract people with backgrounds in business and finance to sit on their boards, unlike American public pension funds, which are stuffed with politicians, cronies and union hacks.

Just as important is their approach to compensation. In order to recruit the best executives, Canadian pension funds have ensured their pay is competitive with Bay Street, Toronto's version of Wall Street. They pay a base salary, annual bonus and long-term performance award (which many pension funds elsewhere do not) to make their employees take a long-term view of investments. Mr Leech of Ontario Teachers made over C3.9m dollar in 2010; 51% of that was a long-term performance award, 36% his annual bonus and only 13% of his base salary. He would doubtless earn more on Wall Street, but this is a huge pay packet by public-pension standards. Anne Stausboll, the boss of CalPERS, the largest American public pension fund, made around 380,000 dollar in the year to June 30th 2011, including a 96,638 dollar bonus.

Such disparity may hinder the Canadian model's spread. Joe Dear, the chief investment officer of CalPERS, has said it is “not politically feasible” to set up this sort of compensation structure. For politicians, not to mention voters, multi-million-dollar salaries are not going to be popular.

Many officials at American public pension funds would not want to copy the model anyway. If a big deal were to go south, they might be sued. But Mr Ambachtsheer says board members also have a fiduciary duty to consider how fees are eroding assets. If they have an option to pay 90% less, they should try to take it.

Another obstacle to the adoption of Canadian ways is scale. A fund needs to be a certain size to buy companies and invest in infrastructure projects, and to swallow the upfront costs of building internal teams. Smaller Canadian funds have been unable to follow suit, for example. Sometimes the large funds will syndicate their deals and give the minnows a chance to take part. In 2009 OMERS won approval to manage assets of smaller pension funds in Canada.

The giants face problems of their own. One, paradoxically, is the growth of assets: CPP, for example, expects to manage C275 billion dollar by 2020 and C1 trillion dollar by 2050. Getting bigger makes it harder for each investment to make a difference to overall returns. “How we can scale our direct investing so it continues to be meaningful in a fund that's doubled its size is a challenge,” says David Denison, CPP's boss.

Another challenge is handling expansion into more volatile emerging markets. To find good deals, funds need people on the ground. But as they become global, they may spread themselves too thinly. The Canadian model assumes that diversification is not as important as deep knowledge: the funds are likely to be better at doing deals in Montreal than Moscow. Some funds have opened foreign offices, but Mr Fyfe of PSP Investments is wary: “These people are going to be sitting there telling you to do a deal, so they're not irrelevant.”