Peer-Multiple Regression

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197

/** Reference calculations for D18-F03 Relative Valuation. */

type Row = Record<string, unknown>;

function numberValue(name: string, value: unknown, options: {positive?: boolean; nonnegative?: boolean} = {}): number {
  if (typeof value !== "number" || !Number.isFinite(value)) throw new TypeError(`${name} must be a finite number`);
  if (options.positive && value <= 0) throw new RangeError(`${name} must be greater than zero`);
  if (options.nonnegative && value < 0) throw new RangeError(`${name} must be nonnegative`);
  return value;
}

function record(name: string, value: unknown): Row {
  if (!value || typeof value !== "object" || Array.isArray(value)) throw new TypeError(`${name} must be an object`);
  return value as Row;
}

function rows(name: string, value: unknown, minimum = 3): Row[] {
  if (!Array.isArray(value) || value.length < minimum) throw new RangeError(`${name} must contain at least ${minimum} rows`);
  const seen = new Set<string>();
  return value.map((raw, index) => {
    const item = record(`${name}[${index}]`, raw);
    if (typeof item.id !== "string" || !item.id.trim()) throw new RangeError(`${name}[${index}].id must be a nonempty string`);
    if (seen.has(item.id)) throw new RangeError(`${name} ids must be unique`);
    seen.add(item.id);
    return item;
  });
}

function median(values: number[]): number {
  const ordered = [...values].sort((a, b) => a - b);
  const middle = Math.floor(ordered.length / 2);
  return ordered.length % 2 ? ordered[middle] : (ordered[middle - 1] + ordered[middle]) / 2;
}

function summary(values: number[]): Record<string, number> {
  const ordered = [...values].sort((a, b) => a - b);
  return {minimum: ordered[0], median: median(ordered), maximum: ordered.at(-1)!, range: ordered.at(-1)! - ordered[0]};
}

function premiumDiscount(implied: number | null, current: unknown, label = "target.current_price"): number | null {
  if (implied === null || current === undefined || current === null) return null;
  return implied / numberValue(label, current, {positive: true}) - 1;
}

function peComparable(inputs: Row): Row {
  const target = record("target", inputs.target);
  const targetEps = numberValue("target.earnings_per_share", target.earnings_per_share, {positive: true});
  const peerMultiples = rows("peers", inputs.peers).map((peer, index) => {
    const price = numberValue(`peers[${index}].price`, peer.price, {positive: true});
    const eps = numberValue(`peers[${index}].earnings_per_share`, peer.earnings_per_share, {positive: true});
    return {id: peer.id, price, earnings_per_share: eps, pe_ratio: price / eps};
  });
  const multipleSummary = summary(peerMultiples.map(row => row.pe_ratio));
  const selected = multipleSummary.median;
  const implied = selected * targetEps;
  return {model: "median-positive-eps-peer-pe", peer_count: peerMultiples.length, peer_multiples: peerMultiples,
    multiple_summary: multipleSummary, selected_multiple: selected, target_earnings_per_share: targetEps,
    implied_price: implied, premium_discount_to_current: premiumDiscount(implied, target.current_price), state: "valuation-complete"};
}

function evEbitda(inputs: Row): Row {
  const target = record("target", inputs.target);
  const ebitda = numberValue("target.ebitda", target.ebitda, {positive: true});
  const shares = numberValue("target.shares_outstanding", target.shares_outstanding, {positive: true});
  const netDebt = numberValue("target.net_debt", target.net_debt ?? 0);
  const preferred = numberValue("target.preferred_equity", target.preferred_equity ?? 0, {nonnegative: true});
  const nci = numberValue("target.noncontrolling_interest", target.noncontrolling_interest ?? 0, {nonnegative: true});
  const nonoperating = numberValue("target.nonoperating_assets", target.nonoperating_assets ?? 0, {nonnegative: true});
  const peerMultiples = rows("peers", inputs.peers).map((peer, index) => {
    const enterpriseValue = numberValue(`peers[${index}].enterprise_value`, peer.enterprise_value, {positive: true});
    const peerEbitda = numberValue(`peers[${index}].ebitda`, peer.ebitda, {positive: true});
    return {id: peer.id, enterprise_value: enterpriseValue, ebitda: peerEbitda, ev_ebitda: enterpriseValue / peerEbitda};
  });
  const multipleSummary = summary(peerMultiples.map(row => row.ev_ebitda));
  const selected = multipleSummary.median;
  const impliedEnterprise = selected * ebitda;
  const impliedEquity = impliedEnterprise - netDebt - preferred - nci + nonoperating;
  const impliedPrice = impliedEquity > 0 ? impliedEquity / shares : null;
  return {model: "median-positive-ebitda-peer-ev-ebitda", peer_count: peerMultiples.length, peer_multiples: peerMultiples,
    multiple_summary: multipleSummary, selected_multiple: selected, target_ebitda: ebitda,
    implied_enterprise_value: impliedEnterprise, equity_bridge: {net_debt: netDebt, preferred_equity: preferred,
      noncontrolling_interest: nci, nonoperating_assets: nonoperating}, implied_equity_value: impliedEquity,
    shares_outstanding: shares, implied_price: impliedPrice,
    premium_discount_to_current: premiumDiscount(impliedPrice, target.current_price),
    state: impliedPrice === null ? "nonpositive-equity-bridge" : "valuation-complete"};
}

function priceToBook(inputs: Row): Row {
  const target = record("target", inputs.target);
  const targetBvps = numberValue("target.book_value_per_share", target.book_value_per_share, {positive: true});
  const peerMultiples = rows("peers", inputs.peers).map((peer, index) => {
    const price = numberValue(`peers[${index}].price`, peer.price, {positive: true});
    const bvps = numberValue(`peers[${index}].book_value_per_share`, peer.book_value_per_share, {positive: true});
    return {id: peer.id, price, book_value_per_share: bvps, price_to_book: price / bvps};
  });
  const multipleSummary = summary(peerMultiples.map(row => row.price_to_book));
  const selected = multipleSummary.median;
  const implied = selected * targetBvps;
  return {model: "median-positive-book-value-peer-pb", peer_count: peerMultiples.length, peer_multiples: peerMultiples,
    multiple_summary: multipleSummary, selected_multiple: selected, target_book_value_per_share: targetBvps,
    implied_price: implied, premium_discount_to_current: premiumDiscount(implied, target.current_price), state: "valuation-complete"};
}

function peg(inputs: Row): Row {
  const target = record("target", inputs.target);
  const forwardPe = numberValue("target.forward_pe", target.forward_pe, {positive: true});
  const growth = numberValue("target.expected_eps_growth_percent", target.expected_eps_growth_percent, {positive: true});
  const peerRatios = rows("peers", inputs.peers).map((peer, index) => {
    const peerPe = numberValue(`peers[${index}].forward_pe`, peer.forward_pe, {positive: true});
    const peerGrowth = numberValue(`peers[${index}].expected_eps_growth_percent`, peer.expected_eps_growth_percent, {positive: true});
    return {id: peer.id, forward_pe: peerPe, expected_eps_growth_percent: peerGrowth, peg_ratio: peerPe / peerGrowth};
  });
  const pegSummary = summary(peerRatios.map(row => row.peg_ratio));
  const targetPeg = forwardPe / growth;
  const impliedPe = pegSummary.median * growth;
  const forwardEps = target.forward_earnings_per_share === undefined || target.forward_earnings_per_share === null
    ? null : numberValue("target.forward_earnings_per_share", target.forward_earnings_per_share, {positive: true});
  const impliedPrice = forwardEps === null ? null : impliedPe * forwardEps;
  return {model: "forward-pe-divided-by-growth-percentage-points", growth_unit: "percentage-points",
    peer_count: peerRatios.length, peer_ratios: peerRatios, peg_summary: pegSummary, target_forward_pe: forwardPe,
    target_growth_percent: growth, target_peg_ratio: targetPeg, selected_peer_peg: pegSummary.median,
    relative_peg: targetPeg / pegSummary.median, implied_forward_pe: impliedPe,
    forward_earnings_per_share: forwardEps, implied_price: impliedPrice,
    premium_discount_to_current: premiumDiscount(impliedPrice, target.current_price),
    state: impliedPrice === null ? "ratio-only" : "valuation-complete"};
}

function solve(matrix: number[][], vector: number[]): number[] {
  const n = vector.length;
  const augmented = matrix.map((row, index) => [...row, vector[index]]);
  for (let column = 0; column < n; column += 1) {
    let pivot = column;
    for (let row = column + 1; row < n; row += 1) if (Math.abs(augmented[row][column]) > Math.abs(augmented[pivot][column])) pivot = row;
    if (Math.abs(augmented[pivot][column]) <= 1e-12) throw new RangeError("regression design matrix is rank deficient");
    [augmented[column], augmented[pivot]] = [augmented[pivot], augmented[column]];
    const scale = augmented[column][column];
    augmented[column] = augmented[column].map(value => value / scale);
    for (let row = 0; row < n; row += 1) {
      if (row === column) continue;
      const factor = augmented[row][column];
      augmented[row] = augmented[row].map((value, index) => value - factor * augmented[column][index]);
    }
  }
  return augmented.map(row => row.at(-1)!);
}

function features(name: string, row: Row): number[] {
  return [1, numberValue(`${name}.expected_growth_percent`, row.expected_growth_percent) / 100,
    numberValue(`${name}.return_on_equity_percent`, row.return_on_equity_percent) / 100,
    numberValue(`${name}.net_debt_to_ebitda`, row.net_debt_to_ebitda)];
}

function regression(inputs: Row): Row {
  const peers = rows("peers", inputs.peers, 6);
  const target = record("target", inputs.target);
  const xRows: number[][] = [];
  const y: number[] = [];
  const multiples: number[] = [];
  peers.forEach((peer, index) => {
    const multiple = numberValue(`peers[${index}].multiple`, peer.multiple, {positive: true});
    xRows.push(features(`peers[${index}]`, peer)); y.push(Math.log(multiple)); multiples.push(multiple);
  });
  const width = xRows[0].length;
  const xtx = Array.from({length: width}, (_, i) => Array.from({length: width}, (_, j) => xRows.reduce((sum, row) => sum + row[i] * row[j], 0)));
  const xty = Array.from({length: width}, (_, i) => xRows.reduce((sum, row, index) => sum + row[i] * y[index], 0));
  const beta = solve(xtx, xty);
  const fittedLogs = xRows.map(row => row.reduce((sum, value, index) => sum + value * beta[index], 0));
  const mean = y.reduce((a, b) => a + b, 0) / y.length;
  const sse = y.reduce((sum, value, index) => sum + (value - fittedLogs[index]) ** 2, 0);
  const sst = y.reduce((sum, value) => sum + (value - mean) ** 2, 0);
  const rSquared = sst === 0 ? (sse === 0 ? 1 : 0) : 1 - sse / sst;
  const fittedPeers = peers.map((peer, index) => ({id: peer.id, observed_multiple: multiples[index],
    fitted_multiple: Math.exp(fittedLogs[index]), log_residual: y[index] - fittedLogs[index]}));
  const targetX = features("target", target);
  const predictedLog = targetX.reduce((sum, value, index) => sum + value * beta[index], 0);
  const impliedMultiple = Math.exp(predictedLog);
  const metric = numberValue("target.valuation_metric", target.valuation_metric, {positive: true});
  const impliedValue = impliedMultiple * metric;
  return {model: "ols-log-multiple-on-growth-roe-and-leverage", peer_count: peers.length,
    feature_order: ["intercept", "expected_growth_decimal", "return_on_equity_decimal", "net_debt_to_ebitda"],
    coefficients: {intercept: beta[0], expected_growth_decimal: beta[1], return_on_equity_decimal: beta[2], net_debt_to_ebitda: beta[3]},
    r_squared_log_space: rSquared, fitted_peers: fittedPeers, target_predicted_log_multiple: predictedLog,
    implied_multiple: impliedMultiple, target_valuation_metric: metric, implied_value: impliedValue,
    premium_discount_to_current: premiumDiscount(impliedValue, target.current_value, "target.current_value"),
    smearing_correction: "not-applied", state: "valuation-complete"};
}

export function calculate(topicId: string, rawInputs: unknown): Row {
  const inputs = record("inputs", rawInputs);
  if (topicId === "D18-F03-A01") return peComparable(inputs);
  if (topicId === "D18-F03-A02") return evEbitda(inputs);
  if (topicId === "D18-F03-A03") return priceToBook(inputs);
  if (topicId === "D18-F03-A04") return peg(inputs);
  if (topicId === "D18-F03-A05") return regression(inputs);
  throw new RangeError(`unsupported topic_id: ${topicId}`);
}