Works of art are incompatible with the traditional notion of investment. Each one is unique and must be judged on its own merits.

As art prices go through the roof, the most hard-boiled money men get excited and start wondering whether the time has come for them to invest in a field where some professionals, better known as art dealers, coin gold.

Judging by the $106.5 million price that Picasso’s "Nude, Green Leaves and Bust," realized at Christie’s New York, paintings can evidently fetch "serious money," as the cliché goes.

Unfortunately, paintings are not a "serious" investment. Indeed, buying art can never be an investment at all, if the word is understood to describe capital expenditure based on calculations that take into account factual data not open to question. Such data do not exist.

Take two experts and you will rarely hear any work of art assessed in identical terms. In Old Master paintings of the 15th and 16th century, mostly unsigned, one eminent scholar will attribute this "Virgin and Child" to So-and-So while the other, equally eminent scholar will say no, it is probably by X, Y, or Z. Or, more subtly, one will see a given picture as entirely from the hand of, say, Lucas Cranach the Elder, and the other will argue that part of it is studio work. And that makes a tremendous difference to the potential value of the picture.

The situation is only slightly different in Impressionist and modern masters. Authorship is rarely at issue (leaving aside authenticity problems, which are fairly rare), but the perception of the importance and quality of a picture may vary. In short, there is rarely universal agreement on the definition of the goods, a prerequisite for any investment.

What makes the concept of investment equally inapplicable to art is the lack of identical units. Consider painting. Every work is unique by virtue of its subject, stylistic handling, brushwork, and, of course, its condition — always a major factor in determining a price.

One Picasso does not equal another. This is why on May 4 estimates for his pictures ranged from $200,000 to $300,000 for "Nature morte au citron," 1947, to $70 million to $90 million for "Nude, Green Leaves and Bust."

Each work must therefore be judged on its own merits. In the case of a painting, this first involves the handling (energetic, subtle, or weak), the brushwork, the appeal of the color scheme, the mood (forceful or dreamy, realistic or Surrealist, sarcastic perhaps). Then come art historical considerations, which include the status of the painter versus other painters from the same school, and the importance of the painting relative to the artist’s entire oeuvre. Since these questions can only be answered on the basis of personal perception, any assessment implies a measure of subjectivity.

Even when those who know something about art and its market broadly agree, the answers they give are never identical. One expert’s vigorous composition may be another’s over-simplified structure, and so forth. Accordingly, one specialist’s estimated presale price may differ from that of a colleague’s. And sometimes, specialists get it all spectacularly wrong anyway.

In May, when the eagerness to buy was greater than ever before, two Picassos missed the lower end of their estimates by roughly 10 percent. At Christie’s, "Le peintre et son modèle," dated November 4, 1964, sold for $10.72 million, and at Sotheby’s, "Buste de femme," painted June 10, 1940, went for $1.59 million. Both pictures were of the most simplistic kind, dashed off in just one day. Estimates for works of that ilk are particularly arbitrary, making the acquisition of each of them a daring gamble for their buyers.

Not surprisingly, two more such Picassos did not sell at all. At Sotheby’s, "Portrait de Sylvette," painted on April 30, 1954, did not elicit a single bid from the room, nor did "Femme sous la lampe," dated April 12, 1962. The estimates were $4 million to $6 million and $3.5 million to $4.5 million, respectively. For those who hoped to sell them, their failure was a nasty shock.

Next