I think there are two difficult issues that puzzle me about macro. First, as much as some people hate the "Wall Street, Main Street" lingo, the question of how financial markets affect the real economy is really important and unsettled. The second question is why labor markets don't adjust to solve problems. Why does a recession show up as unemployment, rather than, say, lower wages at full employment?
I think that the answer to the second question is sociological. But I personally would not spend a whole lot of time on the issue. I may be estranged from my former thesis adviser, Robert Solow, in other ways, but here I agree with him that: (a) you probably need sociology to explain sticky wages and unemployment; but (b) it's not an economist's comparative advantage to pursue it.
To see why it's a sociological issue, think in terms of an example (something Solow also encouraged). Suppose that the job market for third-year law students like my daughter starts to look really grim. Is she going to offer to work for, say, 10 percent less than her competition? Is she going to work as a secretary if she can't find any law-firm jobs? Economic theory suggests she would do either of those things in order to avoid unemployment, but my guess is that neither will happen. I think that the economist with the best feel for this sort of sociology is George Akerlof. But it's not the area I would pursue.
Incidentally, my thesis was an explanation for why nominal prices might be sticky downward. I suggested that if you need to raise your price, you can costlessly drive marginal customers away. But if you lower your price, it costs you advertising money to inform marginal customers of other firms that you are offering a better deal. I mention that because I managed to spend time thinking about what might be an important issue--why are prices sticky, as opposed to solving Euler equations. And look where I am today. Not exactly a tenured professor at an elite university. The moral, in my view, is that it's safer to follow fads. On the other hand, Krugman also ignored the fads to focus on an interesting real-world problem, and he got tenure at Princeton and a Nobel Prize. So you never know.
Anyway, that leaves us with the second issue, of how Wall Street affects Main Street. In my day, a lot of the focus was monetarist, in the sense that people were asking how money could be non-neutral. Keynes had a great story of "animal spirits" of investors and hoarding by savers. But that got jammed into IS-LM, and much was lost.
The Keynesian story is better told in other places. Skidelsky's second volume on Keynes gives a much better flavor of Keynes' views, in my opinion. So that's one book I would recommend. I also think that Charles Kindleberger's ideas in Manias, Panics, and Crashes are worth looking at. I think that Tyler Cowen is channeling Kindleberger when he suggests that one explanation for the recent boom and crash is that new Asian wealth was looking for a place to take risks. That sounds to me like an explanation that Kindleberger might have approved. Finally, George Goodman, aka 'Adam Smith,' does well by Keynes in The Money Game. I'm about to re-read that book, so my opinion could fluctuate.
[update: I re-read the book, and I still recommend it. Don't expect every page to have held up (the book is 40 years old). But it is interesting to compare modern behavioral finance with Goodman's psychological profiles of individual investors and his general focus on the intersection between psychology and finance. The modern practitioners may seem to be more rigorous, but are they? I feel that one learns more reading Goodman than from reading Shiller, not that I am dismissing the latter. My favorite vignette is Goodman's description of a conversatino with Edward Johnson, born at the end of the 19th century and the man who took Fidelity to the top of the mutual fund industry. Johnson comes across as a total airhead. Makes you wonder.]
The bottom line is that the Keynesian view treats financial markets as highly irrational. I think there is much to be said for such a view, even though it presents challenges.
Going in the opposite direction, I think that one should know how modern finance works out the implications of rational markets with objectively known information. You could go back to Modigliani-Miller, to Tobin-Markowitz-Sharpe, and so on.
I think that Fischer Black gives the best description of a pure financial equilibrium, in which I would argue that there is no reason for intermediaries to exist. So Perry Mehrling's biography is essential. Black's own writing ranges from brilliantly clear to hopelessly obscure. On the clear side, I would put, "The Pricing of Commodity Contracts," which explains futures and forward markets. I would also recommend "Bank Funds Management in an Efficient Market," which thinks about what a bank would look like if it understood efficient markets.
Finally, I think that to explain real-world finance one needs asymmetric information. The paper that influenced me was Doug Diamond on delegated monitoring. There are also papers by Stiglitz and Greenwald that are probably important, but they haven't stuck with me as firmly.
The challenge is to weave together these three strands--pure finance, irrationality, and asymmetric information. The pure finance strand is unrealistic, but so is a lot of economics. It still may have value as an "as if" model. For example, index funds outperform most other mutual funds, which is a nice vindication for pure finance (although not necessarily for rationality, if you consider how many people invest in those other mutual funds).
The asymmetric information strand has trouble getting very far. You tend to tackle one or two information problems at a time, when in the real world there are many. But I think you need it in order to develop any theory in which financial intermediaries provide real benefit to the economy, and in which the failure of a financial intermediary imposes real costs. Without asymmetric information, it seems to me that bankruptcy cannot be an interesting event--just liquidate the assets and move on. If new buyers are at a disadvantage in figuring out how to value assets, then bankruptcies are not such trivial events.
I think that without asymmetric information, all assets are liquid, so liquidity preference makes no sense. With asymmetric information, I think you can get liquidity preference. You can get financial intermediation that matters by transforming assets from illiquid to liquid by creating trust, and so forth.
That is why I have trouble saying that all the answers are in Austrian economics, which seems to say that transforming an illiquid asset into a liquid asset is some sort of original sin. My problem is that if we assume no asymmetric information, then everything is liquid. In that world, yes, anything that an intermediary does to introduce risk can only impose costs without benefits.
Once you have information asymmetry, then you cannot say whether the transformations undertaken by intermediaries are harmful or not. They may very well be beneficial, even though they introduce new risks as well.
Finally, you have irrationality. One might like to believe that financial intermediaries profit by providing efficient information-processing services, and that their institutional structures and contract designs are brilliant solutions to problems of incentives in gathering and revealing information. Instead, my guess is that financial institutions evolve to some extent to take advantage of fools. The desire on the part of Asians for dollar-denominated assets has looked foolish to me and to many other economists for a long time. I may be putting words in Tyler's mouth, but I think he at least implies that mortgage securitization was in part a response to the expanding market in Asian fools. Any way you look at it, somebody has been overconfident. The Asians and the skeptical economists can't both be right.
In addition to the readings above, I recommend Liar's Poker for an accurate street-level description of selling and trading securities. There, you will learn something about cultivating markets for fools.