BƯỚC 2. BASELINE RESULTS — TOURISM UNCERTAINTY VÀ DIVIDEND POLICY

Mục tiêu: xác định tourism uncertainty làm firms preserve cash hay maintain/increase dividends as a signal.

PHẦN A. BASELINE AUDIT

Item	Expected	Actual	Pass?
X	country TUI
Y main	DVC/AT
timing	TUI_t → Dividend_t+1
Firm FE	Yes
Year FE	Yes
Country×Year	No main effect
controls	payout determinants
macro	GDPG/Inflation
VCE	country-level
countries/clusters	report

PHẦN B. MAIN RESULT CASES

1. β < 0

Interpretation:

• precautionary cash retention;
• tourism cash-flow uncertainty;
• external financing concern;
• liquidity preservation.

Sample:

Higher tourism uncertainty is associated with lower subsequent cash dividends, consistent with a precautionary payout response in which firms conserve internal liquidity when tourism demand and operating cash flows become less predictable.

2. β > 0

Interpretation:

• dividend signaling;
• investor reassurance;
• agency/catering;
• commitment to dividend stability.

Sample:

Firms increase or maintain payouts when tourism uncertainty rises, consistent with the view that dividends serve as a credibility and reassurance mechanism when external uncertainty is elevated.

3. β insignificant

Do not say TUI does not matter.

Dividend smoothing can make levels sticky.

Check:

• dividend growth;
• dividend payer;
• cut;
• omission;
• repurchases;
• tourism-exposure interaction.

PHẦN C. ECONOMIC MAGNITUDE

If Y = DVC/AT:

$$
Effect_{1SD}

betatimes SD(TUI)
$$

Relative:

$$
Effect_{1SD}/Mean(DVC/AT)
$$

Use percentage-point language carefully.

PHẦN D. DIVIDEND-PAYER RESULT

For:

$$
Payer=1(DVC>0)
$$

use:

• LPM FE as transparent baseline/robustness;
• conditional logit where feasible;
• nonlinear model details separately.

Do not interpret logit coefficient as probability-point change.

Use marginal effects.

PHẦN E. PAYOUT RATIO

Use only positive-earnings sample or exact alternative denominator.

A different sign can occur because uncertainty affects both:

• numerator dividends;
• denominator earnings.

Therefore DVC/AT should remain useful main outcome.

PHẦN F. DIVIDEND CUT / OMISSION

This may be more economically revealing than payout level.

If TUI raises cut probability:

supports precautionary channel.

If firms keep dividends but reduce repurchases:

supports payout smoothing/flexibility hierarchy.

PHẦN G. TOURISM-FIRM VS ALL-FIRM

Compare:

• core tourism firms;
• broad tourism firms;
• all firms in TUI countries.

If effect exists mainly in tourism firms:

stronger tourism-specific interpretation.

PHẦN H. LITERATURE COMPARISON

Important competing evidence:

• Attig et al. — EPU ↑ dividends;
• Choi & Park — EPU ↓ dividends;
• Adra et al. — GPR ↓ repurchases, dividends insignificant;
• recent emerging-market GPR evidence — dividends ↑.

Your result should be framed as resolving which channel dominates for tourism-specific uncertainty.

PHẦN I. HOSPITALITY DIVIDEND CONTEXT

Kilincarslan & Demiralay:

• travel/leisure firms smooth dividends;
• profitability and size increase payment likelihood;
• debt lowers likelihood.

This supports why:

• payout adjustment may be slow;
• debt/liquidity heterogeneity matters.

PHẦN J. MODEL STABILITY TABLE

Model	Controls	FE	β(TUI)	SE	p
1	none	firm+year
2	firm	firm+year
3	+ macro	firm+year
4	+ industry×year	strong

PHẦN K. SAMPLE WRITING

Table X reports the baseline association between country-level Tourism Uncertainty and subsequent corporate dividend policy. The coefficient on TUI is [sign], indicating that firms [reduce/increase/do not materially change] cash dividends following increases in tourism uncertainty. A one-standard-deviation increase in TUI is associated with [magnitude]. The result is consistent with [precautionary/signaling] arguments and should be interpreted against the mixed uncertainty-payout literature, where uncertainty has been shown to both reduce and increase dividends depending on institutional and financing conditions.

PHẦN L. OUTPUT

1. Baseline Audit.
2. Main DVC/AT Table.
3. Payer Table.
4. Payout Ratio Table.
5. Cut/Omission Table.
6. Economic Magnitude Table.
7. Literature Comparison Matrix.
8. Baseline Results draft 1,000–1,500 words.

PHẦN M. CHECKLIST

☐ main sign interpreted using competing theories.
☐ magnitude calculated.
☐ payer marginal effects reported.
☐ positive-earnings payout sample documented.
☐ dividend cut/omission checked.
☐ tourism-only vs broader sample compared.
☐ uncertainty literature mixed evidence acknowledged.
☐ no causal overclaim.

Chuyển sang Bước 3: Robustness.