Merger Arbitrage Model

A risk-arbitrage position model on a single announced cash-and-stock deal: Vantage Analytics Group to acquire Halcyon Systems for 34.00 in cash plus 0.2600 acquirer shares, signed in March 2026. It is not a valuation model with a spread bolted on, and it exists for three mechanics no other archetype carries. The first is that the consideration is collared. The stock leg is fixed between 85.50 and 104.50 on the acquirer price, so inside the collar the exchange ratio is fixed and the deal value moves one for one with the acquirer, while outside it the value is fixed and the ratio moves instead - which means the number of acquirer shares the arb is short is the effective ratio, not the headline 0.2600, and the hedge has to be re-struck once either bound binds. The consideration build therefore runs twice, once at the market price and once across a nine-point grid from 20 per cent down to 20 per cent up, and the grid is what makes the kink visible: deal value pins at 56.23 at the bottom, tracks the acquirer through the middle, and pins again at 61.17 once the upper bound binds. On the shipped case the acquirer trades at 92.40, inside the collar, so the deal is worth 58.02 against a target price of 54.80 - a gross spread of 3.22, or 5.9 per cent, on a 30.4 per cent premium to the unaffected price. The second mechanic is that timing is a hazard curve rather than a date. Five conditions - antitrust, foreign investment, the sector regulator, the shareholder vote and financing - each carry a conditional clearance probability that multiplies down to a joint 90.6 per cent, with the marginal break contribution of each carried alongside so it is visible that antitrust is 4.5 of the 9.4 points of break risk and everything else together is under 5. Nothing can close before month 7; from there a conditional monthly hazard of 18 per cent escalates 35 per cent a month and is forced to one at the month-13 outside date, so the distribution is complete by construction and a check row proves the weights sum to one. The probability-weighted result is 9.20 months, which is the denominator behind every annualised figure in the workbook, and the row of position-open weights it also produces is what carry accrues against - carry in month 12 only counts for the 9 per cent of paths still open in month 12. The carry ledger itself is built on one share: long at 54.80, short 0.2600 acquirer shares worth 24.02, 78.82 of gross exposure against 23.65 of equity posted at a 30 per cent margin and 7.13 borrowed. Four monthly lines run against it - 0.073 of target dividend received, 0.035 of acquirer dividend owed away on the short, 0.012 of stock borrow and 0.030 of financing - netting to a drag of 0.004 a month. That the position is close to carry-neutral is the point, and it is one input away from not being: drop the margin requirement to 25 per cent and financing rises to 0.047 a month, five times the drag, for a position that is only marginally more levered. The ticking fee is deliberately kept out of that ledger, because it is an increment to the merger consideration of 0.14 per share for every month past month 9 paid at closing rather than accrued in cash; weighted across the timing distribution it is worth 0.107, a third of the carry drag and running the other way. The third mechanic is a downside built as a price rather than a haircut. The 44.50 unaffected price is rebased by the 2 per cent the sector has moved since signing, cut by a 10 per cent broken-process overhang, and credited with a 460 million reverse termination fee - 5.48 a share, but payable only on a regulatory failure and so weighted at 65 per cent to 3.56. Break value lands at 42.81, an 11.99 loss on the current price and minus 50.7 per cent of the equity posted, which puts downside to upside at 3.72 times and is why sizing is a sheet of its own. Three branches - completes at 84.6 per cent, a topping bid at 6 per cent worth an 8 per cent uplift, a regulatory break at 9.4 per cent - carry to an expected profit of 2.14 per share, 9.0 per cent on capital and 11.4 per cent annualised, and a bridge underneath decomposes that number exactly rather than by a plug because every outcome value is expressed as a gap to the base case: 3.22 of gross spread, plus 0.11 of ticking fee, less 0.03 of carry, plus 0.28 of topping-bid uplift, less 1.44 of break cost. Read that column and the trade is not a 5.9 per cent spread; it is a 5.9 per cent spread against 1.44 of expected break cost, with the ticking fee and the carry as rounding. The last block inverts the tree. Given a 42.81 downside and a 58.13 upside, the 54.80 market price is already charging a 78.3 per cent completion probability against the model's 90.6, and the breakeven probability at which expected profit is zero is 78.5 per cent - almost exactly the market's number, so the margin of safety is 12.1 points and no more. The same calculation is carried across nine target prices, so the implied probability can be read against the model's at any entry point rather than only at today's. Fund-level sizing then makes the consequence concrete: a 4 per cent weight of an 850 million book is 34 million of capital at 3.33 times gross leverage, 3.07 million of expected profit worth 0.36 per cent of net asset value, against a 17.3 million loss worth 2.04 per cent of it if the deal breaks - 5.6 completed deals to pay for one failure. Five named scenarios and a nine-row check block, all reconciling to zero, close the workbook.