The Second Supplier Is Not Waste. It Is An Option You Forgot To Price

In a piece for forbes.com titled "The Strategic Value Of Optionality," Dr. Aditya V Kashyap, who describes himself as an AI and innovation leader, argues that modern management has spent 50 years waging a largely successful war on slack ^[1][2]. The consequence he identifies is not inefficiency but lost decision rights, and he cites two settlements of that bill: Taiwan's near-monopoly on advanced logic chips and the largest monthly jump in corporate lending on record ^[16][20].

The war was won honestly. Inventories shrank to days of demand, vendor lists consolidated, infrastructure centralized and capital flowed to its highest expected return, and Kashyap concedes each decision was defensible and most were profitable ^[3]. The flaw is structural: optimization assumes tomorrow will resemble the future the system was tuned against ^[4].

His framing is borrowed from finance. An option has value because it grants the right, but not the obligation, to act after uncertainty resolves, and the premium buys the freedom to decide later ^[5]. Optionality, in his definition, is the preservation of decision rights under uncertainty ^[6]. A second supplier preserves the right to redirect supply, unused borrowing capacity preserves the right to liquidity, and a modular stack preserves the right to substitute components after assumptions fail ^[7].

This is where the accounting goes wrong. Redundancy, diversification, modularity and liquidity are mechanisms that create options rather than the options themselves, and none of them produces much under normal conditions, which is exactly why conventional analysis marks them down ^[8]. Their value lies in what they permit, not what they produce ^[9]. Kashyap is explicit that this is not an argument against efficiency: efficiency maximizes performance conditional on assumptions holding, while optionality preserves performance when they fail ^[10]. The deeper threat is irreversibility, such as the proprietary platform that makes migration prohibitive or the capability that takes a decade to rebuild once disbanded ^[11].

Architecture is where the trade shows up on the invoice, because it sets the cost of changing your mind, and tightly coupled systems are cheaper to build and faster to run ^[12]. Modularity, open standards, data portability and clean API boundaries carry a premium paid in integration overhead and forgone optimization ^[13]. He does not recommend flexibility everywhere; duplicating every workload across clouds usually costs more than the option is worth ^[14]. The test is interface by interface: whether the switching cost being created is one the organization could afford to pay ^[15].

The semiconductor case is what happens when nobody pays the premium. A 2021 study by the Semiconductor Industry Association and Boston Consulting Group found that as of 2019, 92% of world fabrication capacity for advanced logic below 10 nanometers sat in Taiwan, leaving 8% everywhere else ^[16][19]. Individually rational decisions collectively left the system with few alternative pathways, and Kashyap describes the public money now spent rebuilding geographic redundancy as the deferred price of options nobody bought ^[17][18].

The exercise event is cleaner still. Federal Reserve economists documented that commercial and industrial lending rose by nearly $480 billion in March 2020, the largest monthly increase in the history of the series, primarily because firms drew down revolving credit lines that had sat unused for years as a fee paid for nothing visible ^[20][21].

Watch the next efficiency drive. Revolver commitment fees and duplicate qualified suppliers are the first line items cut when there is nothing visible to show for them ^[21][8], and the fab subsidies now in flight are the price of the last round of that reasoning ^[18].