Read other letters about this article
I also have trouble with the gambling analogy, although there is some clear truth in it.
The problem with it is, there isn't the total disassociation between the "real" and the "unreal" economy as described here. In sports, a point made on the field has no intrinsic value. Outside of how you bet, there is no way to say "that's a $100 point".
In the world of investments, the betting is on how valuable something is going to be in the future. If I think this high risk loan is likely to pay full value, it's going to be worth $120 next year. If I can get it for $100 today, that might be worth the risk that it's worth $0 in a year.
If people get in the habit of thinking that this $120 next year is less risky than it really is, you have a lot of people buying something for $100 and ending up with nothing. Once this happens a few times, the next guy says, "whoa, wait a minute, I underestimated the risk here, I don't want this anymore and I want to sell it right now for $100". You get enough people trying to sell for $100 and suddenly nobody wants to buy for $100, they want to pay $90, or $80, or $50. Suddenly huge swathes of money just disappear when people own millions of $100 units that are now $50 units.
There's a huge ripple effect from there. Say my bank bought a million units for $100 and the market now thinks they are worth $50. That's $50 million that my bank no longer has. People who invested money in my bank now start to worry if my bank is still a good investment and the stocks and bonds that my bank used to fund these purchases become devalued. The banks and funds that own much of my bank lose value, and so on and so on.
The truth of the betting analogy is that people get carried away with their estimates of what that future value is going to be. They decide that the risks are minimal because the economy will always grow, or because the internet is raining gold on anyone with a smart developer and a smart lawyer, or because oil will never go above $50/barrel, or "they're not making any more land in San Francisco" or whatever.
Also, people bet on what other people are going to bet on. If I think that the future stream of income from amazon.com is going to be a lot higher than most people estimate I'll be willing to bet higher than the next guy will - I'll be willing to buy their stock for more than the next guy and bid up the price, expecting my return on the stock to justify that price.
Now if I think that THE OTHER GUY thinks that the stock price is going up, I may also be willing to pay more for the stock because I think the value of the stock will go up when he outbids me.
So yeah, lots of gambling. But all the gambling is (or should be!) fundamentally rooted in expected future dollar flows. If you lose sight of that, you are in trouble. I think that is the real argument here, that people treat it as though it were as disconnected as betting on sports, and get themselves in trouble.