""" 投资决策 v4.1 — 真实数据修正版 ================================ 基于 009051 易方达中证红利 ETF 联接 真实历史数据 (5.9年年化 8.3%, 18% 波动). 基于沪深 300 全收益指数长期 CAGR ~10%, 25% 波动. 核心修正: - v4 错误: 用 沪深300 (8%) → 中证红利 (8.5%) 假设, 推荐 B 方案 (3000纳+2000红利+2000纯债) - v4.1 修正: 真实数据下, 沪深 300 (10%) 略胜中证红利 (8.3%), 但中证红利低波动是优势 - 新最优: E 方案 (2000纳+2000红利+1000沪深300+2000纯债) 不破产率 89.2% 最高 用户实际操作 (选项 B = 不卖沪深300 + 14,000 建仓中证红利): - 38,000 沪深 300 保留不动 (自然稀释) - 14,000 一次性 或 分批 买中证红利 - 月定投改为: 3000纳+2000红利+2000纯债 (跟 E 方案一致) - 长期 = E 方案 (89.2% 不破产率) """ import numpy as np import json np.random.seed(42) INFLATION = 0.025 N_ACC = 10000 N_RET = 5000 ACC_MONTHS = 120 RET_MONTHS = 1200 # ==== 真实历史参数 (基于 009051 真实数据) ==== ASSETS_ACC = { 'ndx': {'mean': 0.13, 'vol': 0.25}, # 纳指 159696 'hs300': {'mean': 0.10, 'vol': 0.25}, # 沪深 300 (含分红, 散户实际) 'div': {'mean': 0.083, 'vol': 0.18}, # 中证红利 009051 真实 'bond': {'mean': 0.03, 'vol': 0.03}, } ASSETS_RET = { 'ndx': {'mean': 0.08, 'vol': 0.20}, 'hs300': {'mean': 0.06, 'vol': 0.20}, 'div': {'mean': 0.05, 'vol': 0.15}, 'bond': {'mean': 0.03, 'vol': 0.03}, } # ==== 保险 (xlsx 真实) ==== INS_INIT = 1_853_233 STOCK_INIT = 281_246 INS_GROWTH_EARLY = 0.04 INS_GROWTH_LATE = 0.034 STOCK_GROWTH = 0.04 INS_WITHDRAW = 5700 LIVING_COST_50 = 7000 PENSION_63 = 3000 # ==== 5 方案 (A/B/C/D/E) ==== PLANS = { 'A 5000全纳指': {'ndx': 5000, 'hs300': 0, 'div': 0, 'bond': 2000}, 'B 3000纳+2000中证红利': {'ndx': 3000, 'hs300': 0, 'div': 2000, 'bond': 2000}, # v4 推荐 'C 3000纳+2000沪深300': {'ndx': 3000, 'hs300': 2000, 'div': 0, 'bond': 2000}, 'D 2000纳+3000沪深300': {'ndx': 2000, 'hs300': 3000, 'div': 0, 'bond': 2000}, 'E 2000纳+2000红利+1000沪深300 ⭐': {'ndx': 2000, 'hs300': 1000, 'div': 2000, 'bond': 2000}, } def sim_acc(plan): end_ndx, end_hs, end_div, end_bond, max_dd = [], [], [], [], [] for _ in range(N_ACC): ndx, hs, div, bond = 0.0, 0.0, 0.0, 0.0 peak, dd_max = 0, 0 for _ in range(ACC_MONTHS): r_ndx = np.random.normal(ASSETS_ACC['ndx']['mean']/12, ASSETS_ACC['ndx']['vol']/np.sqrt(12)) r_hs = np.random.normal(ASSETS_ACC['hs300']['mean']/12, ASSETS_ACC['hs300']['vol']/np.sqrt(12)) r_div = np.random.normal(ASSETS_ACC['div']['mean']/12, ASSETS_ACC['div']['vol']/np.sqrt(12)) r_b = np.random.normal(ASSETS_ACC['bond']['mean']/12, ASSETS_ACC['bond']['vol']/np.sqrt(12)) ndx = (ndx + plan['ndx']) * (1 + r_ndx) hs = (hs + plan['hs300']) * (1 + r_hs) div = (div + plan['div']) * (1 + r_div) bond = (bond + plan['bond']) * (1 + r_b) tot = ndx + hs + div + bond peak = max(peak, tot) if peak > 0: dd_max = max(dd_max, (peak-tot)/peak) end_ndx.append(ndx); end_hs.append(hs); end_div.append(div); end_bond.append(bond) max_dd.append(dd_max) arr = np.array([a+b+c+d for a,b,c,d in zip(end_ndx, end_hs, end_div, end_bond)]) return { 'p10': float(np.percentile(arr, 10)), 'p25': float(np.percentile(arr, 25)), 'p50': float(np.median(arr)), 'p75': float(np.percentile(arr, 75)), 'p90': float(np.percentile(arr, 90)), 'dd_p50': float(np.median(max_dd)), 'dd_p95': float(np.percentile(max_dd, 95)), 'loss': float(np.mean(arr < 840000) * 100), 'ndx_p50': float(np.median(end_ndx)), 'hs_p50': float(np.median(end_hs)), 'div_p50': float(np.median(end_div)), 'bond_p50': float(np.median(end_bond)), } def sim_ret(plan, acc): succ, end_v, bk_ages = 0, [], [] for _ in range(N_RET): ndx, hs, div, bond = acc['ndx_p50'], acc['hs_p50'], acc['div_p50'], acc['bond_p50'] ins, stock = INS_INIT, STOCK_INIT bankrupt = False for m in range(RET_MONTHS): age = 50 + m/12 ndx *= (1 + np.random.normal(ASSETS_RET['ndx']['mean']/12, ASSETS_RET['ndx']['vol']/np.sqrt(12))) hs *= (1 + np.random.normal(ASSETS_RET['hs300']['mean']/12, ASSETS_RET['hs300']['vol']/np.sqrt(12))) div *= (1 + np.random.normal(ASSETS_RET['div']['mean']/12, ASSETS_RET['div']['vol']/np.sqrt(12))) bond *= (1 + np.random.normal(ASSETS_RET['bond']['mean']/12, ASSETS_RET['bond']['vol']/np.sqrt(12))) ins *= (1 + (INS_GROWTH_EARLY if age < 68 else INS_GROWTH_LATE)/12) stock *= (1 + STOCK_GROWTH/12) ins -= INS_WITHDRAW if ins < 0: ins = 0 cost = LIVING_COST_50 * ((1 + INFLATION) ** (age - 50)) pen = PENSION_63 * ((1 + INFLATION) ** max(0, age-63)) gap = cost - INS_WITHDRAW - pen if gap > 0: if stock >= gap: stock -= gap else: gap -= stock; stock = 0 if bond >= gap: bond -= gap else: gap -= bond; bond = 0 if div >= gap: div -= gap else: gap -= div; div = 0 if hs >= gap: hs -= gap else: gap -= hs; hs = 0 if ndx >= gap: ndx -= gap else: bankrupt = True; bk_ages.append(age); break if not bankrupt: succ += 1 end_v.append(ndx + hs + div + bond + ins + stock) return { 'success': succ/N_RET*100, 'p10': float(np.percentile(end_v, 10)) if end_v else 0, 'p25': float(np.percentile(end_v, 25)) if end_v else 0, 'p50': float(np.median(end_v)) if end_v else 0, 'p75': float(np.percentile(end_v, 75)) if end_v else 0, 'p90': float(np.percentile(end_v, 90)) if end_v else 0, 'mean_bk_age': float(np.mean(bk_ages)) if bk_ages else 150, } print("="*90) print("参数: 纳指 13%/25% 沪深300 10%/25% 中证红利 8.3%/18% 纯债 3%/3%") print("="*90) print("Phase 1: 40-50岁 10年积累期 (10,000次)") print("="*90) print(f"{'方案':<40}{'P10':>10}{'P50':>10}{'P90':>10}{'回撤':>9}{'亏损%':>8}") print("-"*90) accs = {} for name, p in PLANS.items(): r = sim_acc(p) accs[name] = r print(f"{name:<40}{r['p10']/10000:>8.1f}万{r['p50']/10000:>8.1f}万{r['p90']/10000:>8.1f}万{r['dd_p50']*100:>7.1f}%{r['loss']:>7.1f}%") print() print("="*90) print("Phase 2: 50-150岁 100年退休期 (5,000次, 含保险+通胀+养老金)") print("="*90) print(f"{'方案':<40}{'不破产率':>10}{'破产年龄':>10}{'150岁P50':>12}{'150岁P10':>12}") print("-"*90) rets = {} for name, p in PLANS.items(): r = sim_ret(p, accs[name]) rets[name] = r print(f"{name:<40}{r['success']:>8.1f}%{r['mean_bk_age']:>8.1f}岁{r['p50']/10000:>10.1f}万{r['p10']/10000:>10.1f}万") # 保存 output = { 'parameters': { 'inflation': INFLATION, 'assets_acc': ASSETS_ACC, 'assets_ret': ASSETS_RET, 'ins_initial': INS_INIT, 'stock_initial': STOCK_INIT, 'ins_withdrawal_monthly': INS_WITHDRAW, 'pension_at_63': PENSION_63, 'n_sims_acc': N_ACC, 'n_sims_ret': N_RET, }, 'plans': PLANS, 'accumulation': {name: {k: v for k, v in r.items() if k not in ['ndx_p50','hs_p50','div_p50','bond_p50']} for name, r in accs.items()}, 'accumulation_p50': {name: {'ndx': r['ndx_p50'], 'hs300': r['hs_p50'], 'div': r['div_p50'], 'bond': r['bond_p50']} for name, r in accs.items()}, 'retirement': rets, 'cashflow_table': { str(age): { 'cost': LIVING_COST_50 * ((1 + INFLATION) ** (age - 50)) if age >= 50 else 0, 'insurance': INS_WITHDRAW, 'pension': PENSION_63 * ((1 + INFLATION) ** max(0, age-63)) if age >= PENSION_63 else 0, 'gap': (LIVING_COST_50 * ((1 + INFLATION) ** (age - 50)) if age >= 50 else 0) - INS_WITHDRAW - (PENSION_63 * ((1 + INFLATION) ** max(0, age-63)) if age >= PENSION_63 else 0), } for age in [50, 55, 60, 62, 63, 65, 70, 80, 90, 100, 120, 150] }, } with open('sim_v4.json', 'w', encoding='utf-8') as f: json.dump(output, f, ensure_ascii=False, indent=2) print() print("📁 sim_v4.json 已保存 (5 方案)")