I've played this game a few times now and so I thought I'd impart what I've managed to gather.

Oddly enough, the strategies that work well for me in this game are ones that I've adapted from Modern Art: the dynamic of money moving between the players I think is part of the reason.

Where to spend:
Remember that buying machines or warehouses takes money out of the game, so buy sparingly. Buying containers from your opponents' factories or harbors keeps that money available to you for later. They will also be more willing to buy from you because you bought from them. And most importantly, they will have that money with which to bid on containers that you put up for auction, which is how money comes into the game. You want that new money more than you want to hold onto old money!

Early game:
In the early part of the game you should be concentrating on selling small quantities of goods in different singles and pairs (so one orange and one black, for example), for two reasons. First you want to build up capital from new money (money from auctions), and second you want to suss out what color combinations sell for the most money (and to whom). You don't need to keep perfect track, but you want to know what color COMBINATIONS fetch high prices. This is because your opponents will usually only bid on something worth money to them if it also has containers that will pad their slop color (later in the game a poorly chosen color combination will sell for far less than you will be comfortable with, but often for too much for you to justify rejecting the bid).

Early game pricing is usually bargain-basement. In the factory place at least one item for sale at 1, then (if your opponents are not beginners) put whatever complements that container for sale at 2 (if your opponents are beginners just sell it all at 1 because they will not value the action over the money). Same with the harbor and the 2 and 3.

Machines and Warehouses:
When it comes to warehouses, you pretty much NEED two in order to survive in the game. Three can be a wise investment in the midgame if you are doing well at filling your harbor with hot goods. I'd shy away from four or five, though: at that point you're putting yourself in a pretty big hole for goods that more than likely will eventually just stagnate and sell for $2 at the end of the game.

Machines can let you make some early money, but only if you choose the colors carefully. If you are the first to buy a second machine, hope that nobody else buys that color. If you are the last, buy whatever is selling well in combination with your first color. Again, two machines are from my experience very common, and three are occasional. I've never seen anyone buy the fourth. Remember, if containers are sitting in your factory at the end of the game they're worthless, so the money lost on buying a machine needs to be made up somehow. Machines are, however, a good way to get a color you want to sell at auction onto your own ships if they're priced to sell!

Loans:
Don't be afraid of loans. As long as you have a plan for getting them paid off within one or two turns, they can be powerful tools. As soon as you take one, if you don't have a ship almost ready for the foreign island, that needs to become a priority. Use them to: a) let yourself bid for those things you really need, or b) get the correct containers for your own sale. Do not use them for machines or warehouses. Since they are draining your money out of the game over time, you want to repay them with new money, and more new money than you paid.

Mid-late game:
By now you will have some idea of who wants which color combinations and who needs certain colors to complete the set. Because of the balance of the values, most combinations will have someone willing to bid high on them. You can try pricing your goods in the factory or harbor higher if you have managed a monopoly on a color, but in most of the games I've played, the lower prices still win there because SOMEONE is always willing to undercut.

At this stage, the prices in the harbor and factory are far less important than the auctions. In general worry more about getting the right color combinations up for sale than about how much you pay for each container. Also remember that timing is huge: spend the extra $2 or whatever to get everything done NOW while your goods are hot rather than letting someone else get something similar to market.

This is the point where you need to start thinking of money as VP. If you sell goods at auction for $5 you gain 10VP and the other person loses 5VP. If you give $5 to someone else, they will earn 10VP, so make sure the container group you're bidding on will earn you something close to 15VP on them or they are going to be gaining on you. The ideal is to bid as close to the break-even as possible without going under. If you can always sell at a gain and never buy at a loss, you will be very close to winning. You have to guess, but as the buyer you know the hidden value of your container colors, so you can be pretty sure you know how many VP those boxes on the island are.

Now in practice, most auctions you'll have to take a minor deficit as the buyer, but as the seller you will get a minor gain. This is because the new money (VP into the game) goes to the seller. While you probably won't be able keep track of exactly who has how much money, keep an eye on the trends: who buys for really big deficits or really small ones, and so on.

Try to grab containers that the people who are running more deficit want so that they will be your buyers. You can't control who bids the most, but you can grab the missing two colors of the person you know is overbidding consistently and put those up for auction.

Likewise, you can pay more to a seller that you know has been overbidding than you would to someone who is running low VP deficits. However, avoid buying from the same seller over and over! You can't sell to a bad buyer over and over because they will self-eliminate from future auctions, but a shrewd buyer might take you over and over if you keep choosing the same colors.

The more you can control who gets the new money, the better you will do in the game. You can't always be the one getting the new money, but you can be in a position to make sure that either you're blocking too large a windfall from going out or you're getting more big gains than others.

Hope this helps, it's my first stab at a strategy article. Please let me know where I can clarify and expand!