Prakashika Solar Glass can build a plant now for an NPV of Rs 20 crore, or wait a year to learn whether demand is good or weak. At 10%, should it invest now or wait?

Wait: the strategy is worth about Rs 40.9 crore today against Rs 20 crore for investing now. In a year Prakashika builds only if demand is good, earning Rs 90 crore with half probability, Rs 45 crore expected, Rs 40.9 crore today. Committing to build in both states would be worth only Rs 9.1 crore, so the right to walk away is worth about Rs 31.8 crore. The answer flips if waiting cuts the good-state value below about Rs 254 crore.

Because NPV compares investing with doing nothing, and here there is a third choice. Investing now and investing later are mutually exclusive, so the right comparison is between them, not between investing now and zero. Think of buying a house the month before a metro line is either approved or cancelled. Buying now may still be a good deal, but waiting a month lets you buy only if the line is approved. That choice has value, called a real optionThe right, but not the obligation, to take a business decision later, such as expanding, delaying or abandoning a project, once more is known., and the plain NPV of Rs 20 crore ignores it.

Lay out the tree. If the tender goes ahead, the line is worth Rs 300 crore against a cost of Rs 210 crore, a gain of Rs 90 crore. If it does not, building would lose Rs 70 crore, so Prakashika does not build and gets zero. Expected value in a year is half of 90 plus half of zero, Rs 45 crore, worth Rs 40.9 crore today at 10%. That is double the Rs 20 crore from building now.

From avoiding the bad outcome, not from the year's delay. If Prakashika waited but committed to build whatever happened, it would get half of 90 plus half of minus 70, Rs 10 crore in a year, Rs 9.1 crore today, worse than building now. The Rs 31.8 crore gap between waiting and choosing, and waiting and committing, is the value of the right to walk away. Waiting itself has a cost: a line built now is worth Rs 220 crore today, while a line built next year is worth an expected Rs 220 crore a year from now, Rs 200 crore today. The Rs 20 crore difference is the first year of cash flow given up by waiting, and the option is worth paying that price.

Anything that erodes the good state while you wait. Waiting is worth more than Rs 20 crore only while half of the good-state gain, discounted, exceeds 20, which needs a good-state value above about Rs 254 crore. If a rival could build a line during the year and take enough of the tender demand to cut the good-state value from Rs 300 crore to below Rs 254 crore, investing now wins. The same is true if waiting raises the build cost sharply, or if the uncertainty is smaller than it looks: with outcomes of Rs 240 crore and Rs 200 crore, the weak state would not be avoided at all and the option would be worth little.

State the limit. The probabilities and the two outcomes are judgements, and the answer is only as good as them; a sensible board would ask what evidence supports the 50% chance of the tender and how fast a competitor could actually move.