Hermes-Skills/openclaw-imports/stock-analysis/references/dca-screening.md

DCA Screening (阶梯式买入) — High Dividend Candidates

Find stocks suitable for dollar-cost averaging (laddered buying) with high dividend yields. Combines LongPort data with weighted scoring.

Last reviewed: 2026-06-07

---

DCA Scoring Framework

5-dimension weighted score (0-100) optimized for income + value + stability:

Dimension	Weight	Ideal	Logic
Dividend Yield	30%	>10%	Higher = more income while DCA-ing; cap at 20%
PE (TTM)	20%	5-15	Sweet spot: cheap enough for value, not negative
PB Ratio	15%	<1.0	Below book = margin of safety; <0.5 = deep value
5-Day Volatility	15%	<3%	Low vol = smoother DCA entries, less timing risk
YTD Drawdown	20%	-5%~-15%	Pullback = better entry; too deep = fundamental risk

def dca_score(r):
    score = 0
    # Yield (30%): higher = better, cap at 20%
    score += min(r['yield'] / 20.0, 1.0) * 30
    # PE (20%): sweet spot 5-15
    pe = r['pe']
    if pe is None or pe <= 0: score += 5      # negative = risky
    elif pe < 5: score += 15                    # very cheap
    elif pe < 10: score += 20                   # sweet spot
    elif pe < 15: score += 15
    elif pe < 20: score += 10
    else: score += 5
    # PB (15%): lower = better
    pb = r['pb']
    if pb is None: score += 5
    elif pb < 0.5: score += 15                  # deep value
    elif pb < 1.0: score += 12                  # below book
    elif pb < 1.5: score += 8
    else: score += 4
    # Volatility (15%): lower 5d change = better for DCA
    abs_5d = abs(r['five_d'])
    if abs_5d < 1: score += 15
    elif abs_5d < 3: score += 12
    elif abs_5d < 5: score += 8
    else: score += 4
    # YTD dip (20%): negative = better entry
    ytd = r['ytd']
    if ytd < -10: score += 20                   # great entry
    elif ytd < -5: score += 15
    elif ytd < 0: score += 12
    elif ytd < 5: score += 8
    else: score += 4                            # too hot
    return score

LongPort Data Fetching

from longport.openapi import CalcIndex

indexes = [
    CalcIndex.PeTtmRatio,
    CalcIndex.PbRatio,
    CalcIndex.DividendRatioTtm,
    CalcIndex.TotalMarketValue,
    CalcIndex.TurnoverRate,
    CalcIndex.FiveDayChangeRate,
    CalcIndex.YtdChangeRate,
]

# Batch in groups of 10
resp = ctx.calc_indexes(symbols, indexes)
for item in resp:
    dy = float(item.dividend_ratio_ttm) if item.dividend_ratio_ttm else 0
    pe = float(item.pe_ttm_ratio) if item.pe_ttm_ratio else None
    pb = float(item.pb_ratio) if item.pb_ratio else None
    cap = float(item.total_market_value) if item.total_market_value else 0

Candidate Universe (2026-06 snapshot)

🇭🇰 Hong Kong — High Dividend Blue Chips

Code	Name	Yield	PE	PB	Category
3968.HK	招商银行	6.9%	7.1	0.95	银行 (破净)
2318.HK	中国平安	5.4%	6.8	0.89	保险 (破净)
0728.HK	中国电信	6.1%	12.7	0.86	电信 (央企)
1398.HK	工商银行	5.1%	5.8	0.55	银行 (破净)
0883.HK	中海油	5.2%	8.9	1.33	能源
0267.HK	中信股份	4.5%	6.1	0.46	综合 (破净)
3988.HK	中国银行	~5%	~5	~0.5	银行 (破净)
0939.HK	建设银行	~5%	~5	~0.5	银行 (破净)
3416.HK	AGX国指兑	18.6%	N/A	N/A	高息ETF (covered call)
3417.HK	AGX恒科备兑	18.8%	N/A	N/A	高息ETF (covered call)
1088.HK	中国神华	7.0%	16.7	1.82	能源 (煤)

港股 DCA 特点：

• 银行股大面积破净（PB<1），适合长期收息
• 央企分红稳定，但增长有限
• 高息ETF（3416/3417）yield极高但属covered call策略，capital appreciation受限

🇺🇸 US — BDCs (Business Development Companies)

Ticker	Name	Yield	PE	PB	Profile
HTGC.US	Hercules Capital	10.5%	8.5	1.26	科技BDC龙头，YTD-14%
ARCC.US	Ares Capital	10.2%	11.7	0.96	最大BDC，PB<1
MAIN.US	Main Street Capital	5.9%	11.3	1.56	月分红+补充分红
GAIN.US	Gladstone Investment	6.2%	3.3	0.91	小型BDC
GLAD.US	Gladstone Capital	9.3%	10.2	0.90	收入型BDC
HRZN.US	Horizon Technology	25.4%	14.6	0.94	⚠️ 高息但风险高
FSK.US	FS KKR Capital	22.1%	-5.5	0.57	⚠️ PE为负，亏损
PSEC.US	Prospect Capital	23.8%	-6.8	0.37	⚠️ PE为负，分红可持续性存疑

🇺🇸 US — mREITs (Mortgage REITs)

Ticker	Name	Yield	PE	PB	Profile
NLY.US	Annaly Capital	13.2%	7.7	1.07	最大agency mREIT
AGNC.US	AGNC Investment	14.2%	9.0	1.14	agency MBS
ARR.US	Armour Residential	16.8%	9.3	0.91	住宅mREIT
DX.US	Dynex Capital	15.8%	11.1	0.98	多元mREIT

🇺🇸 US — Blue Chip Dividend

Ticker	Name	Yield	PE	Profile
VICI.US	VICI Properties	6.4%	9.8	娱乐REIT，Triple-net
T.US	AT&T	4.9%	7.5	电信，降息受益
MO.US	Altria Group	5.8%	15.0	烟草，稳定现金流
BTI.US	British American Tobacco	5.3%	12.6	国际烟草
XOM.US	Exxon Mobil	~3.5%	~14	能源巨头
O.US	Realty Income	5.3%	50.6	月分红REIT（PE偏高）

🇺🇸 US — High Yield ETFs

Ticker	Name	Yield	Strategy
JEPI.US	JPMorgan Equity Premium Income	8.3%	ELN + stock selection
JEPQ.US	JPMorgan NASDAQ Equity Premium	10.4%	NASDAQ版JEPI
SPYI.US	Neos S&P 500 High Income	11.9%	S&P 500 covered call
QYLD.US	Global X NASDAQ 100 CC	11.7%	ATM calls on QQQ
SVOL.US	Simplify Volatility Premium	22.3%	波动率溢价

Pitfalls

• ⚠️ Ultra-high yield (>20%) = red flag: HRZN/FSK/PSEC/SVOL yield 20%+ but PE is negative — recent losses. Dividend sustainability at risk. Always check PE > 0 before recommending.
• ⚠️ mREITs are rate-sensitive: NLY/AGNC/DX/ARR depend on net interest margin. Fed rate cuts = tailwind; rate hikes = headwind. Best DCA during rate-cutting cycles.
• ⚠️ Covered call ETFs cap upside: QYLD/JEPQ/SPYI generate income by selling calls — total return may lag underlying index in bull markets. DCA works better in range-bound markets.
• ⚠️ HK bank PB<1 is structural: Chinese bank "破净" has persisted for years — it reflects real estate risk, not necessarily a bargain. Still fine for dividend income but don't expect PB reversion.
• ⚠️ BDC vs mREIT: BDCs (HTGC/ARCC/MAIN) have more diversified income sources and typically more stable dividends than mREITs. Prefer BDCs for conservative DCA.
• Seasonality: US ex-dividend dates cluster around Feb/May/Aug/Nov for quarterly payers. Plan DCA entries to capture dividends.

DCA Strategy Tips

1. 3-5 price levels: Set 3-5 buy levels below current price, spaced 5-10% apart
2. Equal dollar amounts: Invest same $ amount at each level (not equal shares)
3. Dividend reinvestment: DRIP accelerates compounding during DCA accumulation
4. Sector diversification: Mix REIT + BDC + utility + telecom — don't over-concentrate
5. HK + US mix: HK for value (low PE/PB), US for yield (higher dividend rates)