The first step in a Graham and Dodd valuation is to calculate the asset value of the company. For many traditional value investors, this has been essentially the only step. But even within this restricted approach, the investor has to make judgments about the reliability of the information and the strategic situation of both firms and industry in order to make accurate
calculations of the value of the assets.
The strategic judgments concern the future economic viability of the industry (or industries) in which the firm operates. If the industry is in serious decline, then the asset values of the company should be estimated based on what they will bring in liquidation. Since there will be no market for capital goods tailored to specific requirements of the industry, they will basically be sold for scrap. On the other hand, if the industry is stable or growing, then the assets in use will need to be reproduced as they wear out. These assets should be valued at their reproduction cost.
The reliability issue is largely a question of how far down the balance sheet the investor chooses to go. At one extreme, Benjamin Graham considered only current assets (cash, accounts receivable, inventory, etc.) in his valuation. These can be determined within a narrow margin of error using either a liquidation or reproduction cost basis. With the virtual disappearance of Graham's net-net stocks (market price less than current assets minus liabilities), contemporary value investors have moved down the balance sheet to include in their valuations plant, property, and equipment and even intangibles such as product portfolios, customer relationships, brand images, and trained employees. The error band here is certainly wider, and valuing these assets, especially the intangibles, has required
both skill and imagination. Obviously, this effort is worthwhile only for firms operating in viable industries; intangibles are worthless if the industry disappears.
Source: Benjamin Graham's Net-Net Stock Strategy/ Harriman House