Leaving a market: what Target Canada and Uber show about fixed costs

On 15 January 2015 Target announced it would stop operating stores in Canada, where it had 133 stores and about 17,600 employees. Its chief executive said management had been unable to find a realistic scenario that would get Target Canada to profitability until at least 2021 (Target Corporation, 2015). In fiscal 2012 the Canadian segment had still reported no sales at all (Target Corporation, 2015). This post uses Target's filings, and Uber's 2019 prospectus for its exit from Southeast Asia, to ask how to tell a market that cannot carry your fixed costs, how much of an exit's cost was already spent, and what a firm keeps when it leaves.

How can you tell that a market cannot carry your fixed costs?

Set the margin your sales earn beside the costs that do not fall when sales do. If the margin from the whole market cannot cover those costs at any volume you can plausibly reach, more effort inside that market will not repair it. Target's Canadian segment lets you do the sum.

Source: Target, Form 10-K for fiscal 2014, Note 6 (Target Corporation, 2015).

Sales grew 44 per cent in 2014 and the loss narrowed by US$149m, yet the segment still lost about 46 cents on every dollar it sold. Treat SG&A, depreciation and interest as fixed. That is crude: store wages move with trading hours, and interest is a financing choice. At the 2014 gross margin of 19.0 per cent, those US$1,230m would need sales of about US$6.5bn, 3.4 times what the stores sold. This is my arithmetic, not Target's, and a bigger chain would earn a better margin through buying scale, so read the multiple as the size of the gap, not as a break-even point. The 10-K gives no forecast: the 2021 date is management's, from the press release, and no outsider can test it.

What does leaving cost, and how much of it was already spent?

Most of the headline cost of Target's exit was the write-off of money spent earlier. Of US$5,105m in pretax exit costs, US$4,766m, or 93 per cent, was the impairment recorded when Target deconsolidated its Canadian subsidiaries (Target Corporation, 2015).

The exit did not create that loss. It forced Target to record it: according to the 10-K, the subsidiaries' estimated liabilities exceeded the value of assets available to creditors, so the investment was left with a fair value of about zero (Target Corporation, 2015). Accounting decides when a loss appears, and Harrigan notes that capitalised assets can deter exit because leaving before they are depreciated would show a reported loss (Harrigan, 1981, p. 307). My reading: a manager judged on this year's profit has a reason to wait, while an owner choosing where the next dong goes has none, because the money is gone either way. That is the sunk-cost trap in accounting form.

The cash that depends on the decision is smaller and was announced in advance. Target guided to US$500m to US$600m of cash costs, mostly in fiscal 2015 or later, and expected the decision to raise earnings from fiscal 2015 and cash flow from fiscal 2016 (Target Corporation, 2015). In 2014 the segment lost US$869m before tax and spent US$228m on capital (Target Corporation, 2015). The two are not like for like, because the loss includes US$248m of depreciation that is not cash, so they indicate scale, not a payback period. A year later Target's next 10-K said the exit had raised after-tax cash flows from 2015, which is the company's own statement (Target Corporation, 2016).

Does leaving a market mean abandoning what you built?

Not always. Uber's exit from Southeast Asia turned an operating business into a minority stake. On 25 March 2018 Grab hired Uber's employees and acquired its assets in the region, including rider, driver and eater contracts. Uber received Grab Series G preferred shares, first recorded at a fair value of US$2.2bn, and booked a US$2.3bn gain; the net assets it handed over were "not material" (Uber Technologies, Inc., 2019, pp. F-80–F-81).

The stake is a financial asset, not an operation, and it came with strings. The prospectus puts Uber's holding at 30.0 per cent at closing and estimates 23.2 per cent at 31 December 2018. Uber is contractually restricted from competing with Grab in certain aspects of its business in the region until the later of March 2023 or one year after it disposes of all its Grab interests. It may put the shares back to Grab for cash from 25 March 2023, subject to conditions that include Grab not having listed, though the filing calls collection of the contractual interest uncertain. It also recorded US$116m of tax expense on the divestiture in 2018 (Uber Technologies, Inc., 2019).

Why did Uber leave? The prospectus passages I read do not say. Reuters, reporting the announcement, said Uber had invested US$700m in the region and that discounts and promotions across Asian ride-hailing had cut margins and pushed towards consolidation. It quoted the chief executive's note to staff: "One of the potential dangers of our global strategy is that we take on too many battles across too many fronts and with too many competitors" (Reuters, 2018). The logic differs from Target's: Target said its market would not reach profit, while the note names over-extension as a risk of Uber's global strategy as a whole. Nothing I read says Uber's business could not cover its fixed costs.

Berry's study of 190 US firms over 1981 to 2000 treats divestment as a response to struggling operations and also to better uses for resources in other markets (Berry, 2010). I read only the abstract, and it concerns home-country operations, so Uber is a fit by analogy, not a test.

Vietnamese readers saw the cost on the ground. Uber closed in Vietnam on 8 April 2018, two weeks after the announcement, and Viet Nam News reported that the notice seemed too short for drivers and staff (Viet Nam News, 2018). On 16 April 2018 the Vietnam Competition Authority (VCA), part of the Ministry of Industry and Trade, opened a preliminary investigation into the deal. In June 2019 the Vietnam Competition Council rejected the VCA's conclusion that Grab had breached the law and ruled that the acquisition was not an economic concentration (Viet Nam News, 2019). Selling to your closest rival can bring a regulator into the exit.

Why do firms stay too long in a market that loses money?

Because leaving has barriers that are separate from the market's economics. Porter's 1976 article defines exit barriers as factors that inhibit companies from divesting unprofitable businesses, and traces them to the economic characteristics of the business, corporate strategy choices, information problems and managers' conflicting goals when deciding to divest (Porter, 1976). I read only the abstract.

Harrigan tested the idea on 61 firms in eight declining businesses between 1965 and 1978 (Harrigan, 1981, p. 309). She counted the cost of closing a plant and the lack of a resale market as economic barriers, and image, customer obligations and shared facilities as strategic ones (Harrigan, 1981, pp. 307–308). She also expected that where barriers are high, competition among the "trapped" firms could grow more volatile, because their cost structures encourage price cutting to fill plants (Harrigan, 1981, p. 308). Average likelihood of exit was 23.3 per cent where the business was of high strategic importance and 58.1 per cent where it was of low importance (Harrigan, 1981, p. 321). Her advice is to plan the exit when you enter (Harrigan, 1981, p. 322).

How can a small business run the same test on one branch?

Write the break-even, the exit cost and the plan to reach break-even side by side, and fix a date to decide. The example is invented: Company F, a cosmetics distributor, and its branch in a provincial city, with figures in VND million per month.

The branch sells 220 at a 30 per cent gross margin, so gross profit is 66. Fixed costs are 90 (rent 40, staff 30, other 20), so it loses 24 a month. Break-even sales are 90 ÷ 0.30 = 300, 36 per cent above today. If the city's category sells 2,200 a month, the branch holds 10 per cent and needs 13.6 per cent. The lease has 30 months left, and leaving early costs six months' rent (240) plus two months' pay for four staff (60): 300, once.

If sales stay flat, the 300 exit cost equals 12.5 months of losses (300 ÷ 24). If the plan works, option B reaches break-even in month 9 after 108 of cumulative losses. So the question is not "is it losing?" but "what would have to be true for sales to reach 300, and what would show it early?" Fix a tripwire now: if sales are below 250 in month 4, close. Waiting those four months costs at most 96.

Option D shows the difference between leaving and abandoning: the city still contributes, with no rent or payroll, though the dealer figure is an assumption. The test fits a branch, city or channel with measurable sales and fixed costs. It fits poorly where the unit feeds the rest of the business through shared stock, brand or key customers, which Harrigan lists among strategic barriers (Harrigan, 1981, pp. 308–309).

Going deeper: where does the exit-barrier idea come from, and what can these sources not show?

The concept starts with Porter's 1976 article and with Caves and Porter's chapter on barriers to exit, published the same year, which I know only through Harrigan's citation (Harrigan, 1981). Harrigan's results differ from that chapter on one point: Caves and Porter found no statistically significant pattern for manufacturing-related barriers, while her models indicate that durable, specific assets could be a highly significant influence in lowering the chance of early exit (Harrigan, 1981, p. 322). That is a disagreement between two early studies, not a settled finding.

Limits of the academic sources. Harrigan's eight businesses, from percolator coffee makers to baby foods, were declining product lines, not retail or services (Harrigan, 1981, p. 309). Only slightly more than half of the 36 competitors she interviewed would supply disguised financial figures, so she built a relative scale from interviews (Harrigan, 1981, p. 311). Berry's firms divested home-country operations.

Limits of the cases. Target is one case, and its US$5.1bn is mostly an accounting write-off, not a cash loss. I did not read the court monitor's reports, so I cannot say what the leases and stores fetched; the 2016 10-K says only that net assets were valued at the liquidation price received (Target Corporation, 2016). Uber's US$2.2bn is a fair value at the time, for preferred shares in a company that had not listed. I did not read later filings, so "kept value" means value recorded at exit, not value realised. Neither case shows that the freed resources earned more elsewhere, which is the claim behind "portfolio discipline".

The library files Porter's generic strategies under this topic. A low-price position leans on volume to carry fixed costs; that is my reading, and the Target filings do not say whether the Canadian stores followed one. Other exits: Tesco leaving the US and Unilever separating its ice cream.

What did I read? On 3 November 2026 I read in full Harrigan (1981), from a scanned copy I OCR'd (I used only its text and Table 7, since the regression tables did not scan cleanly), Target's press release, the Reuters report and the two Viet Nam News items. Of Target's two 10-Ks I read the Canadian passages, and of Uber's S-1 the Southeast Asia passages, not the whole documents. Porter (1976) and Berry (2010) I read in abstract only.