ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
Turkey Output-Gap Series -- cross-method estimates
2
3
TitleEstimating the Output Gap in a High-Inflation Emerging Market
4
AuthorEmrehan Aktug (Sabanci University)
5
SeriesSmoothed (two-sided) output-gap estimates for Turkey from eighteen estimators.
6
UnitsPercentage points of potential GDP. Positive = output above potential.
7
FrequencyQuarterly.
8
Full sample2005Q1-2025Q4 (84 quarters).
9
Survey sample2013Q1-2025Q4 (52 quarters); matches the companion natural-rate study.
10
Blank cellsIndicate the method is not defined at that date (e.g., Baxter-King drops 12 quarters at each end).
11
Data sourceCBRT EVDS (rates, FX, REER, inflation expectations, capacity utilization) and TURKSTAT (real GDP, CPI).
12
CitationAktug, E. (2026). "Estimating the Output Gap in a High-Inflation Emerging Market."
13
Generated2026-05-29
14
Update stepsRefresh EVDS_DatasetLONGER.xlsx, run code/run_all.py, then run code/export_gap_series.py.
15
16
17
Series (column header)FamilyDescription
18
HP two-sided (lam=1600)Univariate filterHodrick-Prescott filter, two-sided, smoothing parameter lambda=1600.
19
HP one-sided (lam=1600)Univariate filterOne-sided (recursive, real-time) HP filter; end-of-sample reading at each date.
20
HP boosted (Phillips-Shi 2021)Univariate filterBoosted HP filter (Phillips & Shi, 2021); iterated to remove residual trend.
21
Hamilton (h=8,p=4)Univariate filterHamilton (2018) regression filter; horizon h=8, p=4 lags.
22
CF band-pass (6-32q)Univariate filterChristiano-Fitzgerald asymmetric band-pass, 6-32 quarter cycles.
23
BK band-pass (6-32q,K=12)Univariate filterBaxter-King symmetric band-pass; drops K=12 observations at each sample end (hence blanks).
24
Beveridge-NelsonUnivariate filterBeveridge-Nelson decomposition; AR(p) on output growth with BIC lag selection.
25
BN-KMW (refined, AR(12))Univariate filterRefined Beveridge-Nelson (Kamber, Morley & Wong, 2018); high-order AR(12).
26
Linear detrendUnivariate filterDeterministic linear trend; cycle is the OLS residual.
27
Quadratic detrendUnivariate filterDeterministic quadratic trend; cycle is the OLS residual.
28
Piecewise-linearUnivariate filterBroken-line trend with breaks at 2018Q3, 2020Q1, 2021Q4.
29
Mueller-Watson (q=4)Univariate filterLow-frequency projection on a linear trend plus q=4 cosine basis functions.
30
Kuttner closed (regimes)Structural state-spaceClosed-economy Phillips-curve UC model (Kuttner, 1994) with regime variance scaling.
31
HLW open (regimes)Structural state-spaceOpen-economy Holston-Laubach-Williams model (adds REER and FX terms), regime scaling.
32
HLW real-rate (regimes)Structural state-spaceHLW open plus the ex-ante real-rate IS channel; conditions on the monetary-policy stance.
33
HLW-SV (EWMA stochastic volatility)
Structural (diagnostic)HLW with a data-driven EWMA stochastic-volatility path; reported as a diagnostic.
34
Bayesian Kuttner (open, MH)Structural state-spaceBayesian open-economy Kuttner via adaptive Metropolis-Hastings; posterior-median gap.
35
Capacity-utilization gapSurvey-basedHP-detrended manufacturing capacity-utilization rate, rescaled to output-gap volatility.
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