_Accounting was designed to count factories and inventory. It struggles with everything else._ --- ## Precision without accuracy Financial statements are precise to the penny. They're also increasingly wrong about what matters. In the 1950s, earnings explained around 90% of stock price movements. Today, that figure has collapsed to roughly 50%. The market is telling us something: accounting no longer captures what creates value. The reason is structural. Accounting emerged from an industrial economy where value resided in tangible assets — factories, machinery, inventory. You could count these things, depreciate them on a schedule, and the numbers meant something. That economy is largely gone. Value now comes from intangibles: patents, brands, customer relationships, software, organisational capabilities. Accounting doesn't know how to handle them. A widening gap opens between reported numbers and economic reality. --- ## The phantom of product cost The distortion runs deeper than missing intangibles. Cost accounting itself warps reality. Consider "product cost" — one of the most common numbers in business. Managers use it to decide pricing, product mix, make-vs-buy decisions. It feels objective, like a fact about the world. It isn't. You never actually pay money to a product. You pay wages, rent, materials, utilities. These get allocated to products through formulas that are ultimately arbitrary. Change the allocation method and product costs change — without any change in economic reality. A product that looks profitable under one allocation becomes unprofitable under another. This matters because cost accounting drives behaviour. Managers optimise for metrics that don't reflect value creation. They chase cost reductions that look good locally but destroy throughput globally. The classic example is keeping machines busy. Cost accounting rewards high utilisation — spreading fixed costs across more units lowers the "cost per unit." So managers keep everything running, even when there's no demand. Inventory piles up. Cash gets trapped. The constraint sits idle while non-constraints overproduce. The numbers improve while the business gets worse. --- ## Throughput as an alternative Goldratt proposed a different framework, stripped down to three measurements. **Throughput** is money generated through sales — selling price minus truly variable costs, primarily materials. This is the rate at which the system generates cash. **Inventory** is money the system has invested in things it intends to sell. This is cash trapped in the system. **Operating expense** is money spent converting inventory into throughput. This is the ongoing cost of running the system. No product costs, no allocations, no phantom margins. This shift matters. Value is realised at the moment of sale, not during production. "Adding value" to a product before it sells is an illusion — you've added cost, not value. Value only exists when a customer pays. This reframes decisions. Instead of asking "what's the product cost?" you ask: what is the impact on throughput, inventory, and operating expense? The first question invites allocation games. The second forces clarity about cash. --- ## Following the cash Finance sharpens this further. Accounting profits can be manipulated through provisions, depreciation schedules, revenue recognition, and stock option expensing. Two companies with identical economic performance can report very different profits depending on their accounting choices. Cash is harder to fake. You either have it or you don't. This is why sophisticated investors focus on cash flow statements rather than income statements. Cash flows avoid the distortions of accrual accounting. They show real economic capacity: the ability to invest, repay debt, and return money to shareholders. For managers, this means tracking cash generation alongside accounting profits. A business can be profitable and broke simultaneously if cash is trapped in working capital. [[Notes/Profitable And Broke|Profitable and Broke]] makes this concrete — working capital constraints kill growing businesses that look healthy on the income statement. --- ## Looking forward, not backward Accounting characterises what already happened: historical costs, past revenues, accrued expenses. Finance looks forward. The value of a business is the present value of its future cash flows. History matters only insofar as it helps predict what comes next. This changes what you should measure. Accounting asks what your margins were last quarter. Finance asks what margins you can sustain, and for how long. Accounting reports that you earned a certain profit. Finance asks whether you beat your cost of capital. A profitable business that earns less than its cost of capital is destroying value, regardless of what the income statement shows. Three imperatives follow: beat your cost of capital, sustain the gap for as long as possible, and reinvest at high rates. These connect directly to strategy — competitive advantage, barriers to entry, growth opportunities. --- ## Constraints, not averages Cost accounting optimises averages. Throughput thinking focuses on constraints. The difference matters because systems don't behave like averages. In any dependent system with variability, a small number of variables determines almost everything. The constraint — the weakest link — sets the pace for the whole system. Improving non-constraints doesn't improve the system. It creates excess capacity that can't be used, inventory that can't be sold, costs that don't generate returns. This is why [[Notes/Hidden Bottleneck|Hidden Bottleneck]] matters. Often the real constraint is decision speed, not a machine or a team. Weeks disappear waiting for approvals while the operational machine sits idle. Cost accounting won't reveal this. It measures utilisation of resources, not throughput of value. Constraint thinking asks different questions. Where is the bottleneck? Are we exploiting it fully? Is everything else subordinated to serving it? These cut through the noise of local metrics to focus on what actually limits performance. --- ## Better measures If the default numbers mislead, what should you track instead? Start with cash. Track cash generation, cash conversion cycle, free cash flow — these tell you whether the business can fund itself and grow. Then evaluate decisions by their impact on system-wide throughput rather than phantom product margins. Does this increase the rate at which you generate cash through sales? Look forward, not backward. By the time earnings are reported, the information is stale. Focus on metrics that precede financial results: customer adoption, pipeline velocity, retention rates, constraint utilisation. Two more shifts matter. First, profitability alone isn't enough — a business earning 8% on capital when capital costs 10% is destroying value regardless of what the income statement shows. Second, measure what's happening at the bottleneck rather than chasing average utilisation everywhere. Protect constraint uptime. Accept that non-constraints should have spare capacity — keeping them busy creates queues and destroys flow. --- ## The test Pull up your last major decision that used "product cost" or "margin by product line." Trace the allocations. Ask: if we changed the allocation method, would the decision change? If yes, you're optimising for arithmetic, not economics. Then look at your cash. Calculate your cash conversion cycle — days from paying suppliers to collecting from customers. If it's lengthening while revenue grows, you're funding growth with working capital, and [[Notes/Profitable And Broke|profitable businesses die this way]]. Finally, find your constraint. Not the resource that's busiest, but the one that limits system throughput. Measure its utilisation, its output per hour, the queue waiting for it. Everything else is noise. The numbers your accounting system produces were designed for a different economy. They're precise, auditable, and increasingly irrelevant. Build your own measures — throughput, cash, constraints — and make decisions from those. --- **This essay synthesises ideas from:** - [[Notes/From Data To Information|From Data to Information]] - [[Notes/Hidden Bottleneck|Hidden Bottleneck]] - [[Notes/Profitable And Broke|Profitable and Broke]] **See also:** [[Ideas/Theory of Constraints|Theory of Constraints]] · [[Ideas/Goodhart's Law|Goodhart's Law]] · [[Ideas/McNamara Fallacy|McNamara Fallacy]] · [[Book Summaries/The End of Accounting|The End of Accounting]] · [[Book Summaries/How Finance Works|How Finance Works]] · [[Book Summaries/The Haystack Syndrome|The Haystack Syndrome]] **See also:** for the complete collection of Notes.