I used to joke that 95% of investing success consists of knowing what not to do. It's rapidly approaching 99%, and it's no longer a joke.

Look at exchange-traded funds, arguably the most investor-friendly innovation of the past few decades. In their purest form as market-tracking index funds, they offer broad diversification at high tax efficiency and amazingly low cost.

In their latest incarnations, ETFs often offer exposure to narrow, risky assets, high tax bills and bloated fees. Many ETFs have become publicly traded gambling chips. Your job as an investor is to resist temptation and to sort through these new offerings without mercy.

At least three ETF sponsors have filed to launch funds that would go up or down based on which political party wins upcoming elections. More than 100 proposed ETFs are tied to the real-time success of National Hockey League teams over the course of a season-including 32 that would double the daily change in a composite statistical measure of a team's performance.

Since 2019, when the SEC adopted a rule to streamline the launch of new ETFs, asset managers have rolled out hundreds of specialized funds. Among them are ETFs that track cryptocurrencies or single commodities, "buffer" the sharpest market downturns, double the daily returns of single stocks, use options trading to generate double-digit or even triple-digit reported yields, and invest in technologies that might be used by UFOs .

No wonder the SEC recently asked for public feedback on whether it should change how it reviews offerings for what the agency calls "novel ETFs," especially because many such funds propose investing in assets that might not even meet the traditional definition of securities.

"Once you open the door for things of this nature, it's hard to say what's right and what's wrong-or what's good and what's bad," says Dave Mazza, chief executive of Roundhill Investments, which manages about $36 billion in 55 ETFs.

To be fair, even the most unusual ETFs have some investment use. If you run a small business that relies heavily on federal contracts or approvals, a prediction-market ETF that does well when one political party wins might be a useful hedge. If you're the head of sales for a big company that sponsors an NHL team, a hockey ETF could help you manage the risks of a losing season or a star player gone rogue. And such funds are unlikely to move in sync with stocks overall, potentially providing another source of return in a market decline.

But most people can already make similar bets outside a brokerage account. If you want to gamble on the Anaheim Ducks, do you need an ETF? Just invite your buddies over, put the game on the TV and toss $20 on the coffee table-or open any sports-betting app on your phone.

Novel ETFs, on average, are also expensive. Of course, most ETF assets are in funds that cost next to nothing, such as iShares Core S&P 500, which holds more than $800 billion, and Vanguard Morningstar Total Stock Market, with $2.34 trillion in assets. They charge a paltry 0.03%. But 49% of all ETFs charge at least 0.5% in annual expenses, up from 42% in 2019, according to Morningstar. Over the same period, boring old mutual funds have gotten cheaper.

All this gets at the heart of a perennial problem: To be an investor, you have to know the difference between investing and gambling. It's OK to gamble a little from time to time, if-but only if-you know you're gambling and don't think you're investing.

"Many investors, more likely men than women, have a portfolio they trade, and it's not totally wrong or inappropriate to have that," says Deborah Fuhr, founder of ETFGI, a research firm based in London. "But it's very important not to think of it as an 'I'm-going-to-retire-by-using-only-this' portfolio."

When a gamble gets packaged into an investment vehicle like an ETF, it might not feel like a gamble. It could even feel "safe." And making this sort of bet inside your brokerage account could contaminate your thinking.

"Investors are under a kind of two-pronged attack," says Dave Nadig, president and director of research at ETF.com. "The culture is pushing everyone toward speculation and gambling at the same time as the financial markets are getting deregulated. It's going to be very hard to put these genies back in the bottle."

Some of the novel ETFs, with their hyperconcentrated portfolios, are a throwback to more than a century ago.

In the 1920s, before federal investment regulations were created, funds could invest pretty much however they pleased. In 1927, 5% of all funds had the bulk of their assets in only a single investment; in 1929, more than an eighth of all funds had at least 25% of their total assets in a single holding.

That wasn't the best idea. The Swedish American Investment Corp. put most of its assets under the sway of Ivar Kreuger, an aggressive dealmaker who later died in an apparent suicide after his business empire crumbled. The fund was dissolved in the 1930s.

Not all novel ETFs will go bad, but you need to treat them with caution. Here's a list of simple questions to ask before you invest in any new ETF:

Is this fund the easiest and cheapest way to accomplish my goal? How does it fit with the rest of my portfolio?

Does this fund serve an investment purpose, or is it a gamble? If I'm gambling, how will I segregate this bet from my long-term accounts?

Finally and most important: How will I limit my losses-and how badly would I regret them?