ๆญฏ๋ฏธ๊ตญ๊ฒฝ๊ธฐํ‘œ์ง€.PDF

Size: px
Start display at page:

Download "ๆญฏ๋ฏธ๊ตญ๊ฒฝ๊ธฐํ‘œ์ง€.PDF"

Transcription

1 2001 3

2 LG

3 < > < > I II III IV V 51 < >53

4 < > <I-1> 3 <III-1> (1970~1999) 25 <III-2> <IV-1> G-7 ( ) 44 < I-1> 4 < II-1> (1985=100) 8 < II-2> 9 < II-3> 13 < II-4> S&P500 (PER) 16 < II-5> 20 < III-1> G-7 26 < III-2> GDP (1%) 29 < III-3> GDP (1%) 35 < IV-1> 43 < IV-2> 43 < IV-3> 45 < IV-4> (1%),,, 46 < IV-5> / / 49 < IV-6> / / 50 Box I-1 (FRB) 2 Box II-1 (Rcssion Probabiliy) 22 Box III-1 G-7/KR-SVAR 30 Box III-2 US/KR-SVAR 37

5 < > ,000 NASDAQ 1 2,000 10,000 (Tch Sock) 2 1/4 53% : F S Mishikin LG i

6 53% Gordon 2 (horical valu) S&P , NASDAQ 1,528 30% 30% 320 FRBFOMC ii

7 iii 2002 (corrcion) ( ) % U / % 20022/4 30% 20012/4 3 01% / /4 U 90

8 GDP1% 4 GDP 06% 20% GDP1% G7 (99 ) GDP( )07% GDP 50% (GDP 1% ) : LG iv

9 v 5% 3~4% 3% Coningncy Plan,

10 / /4 11% 1/4 (FRB) ~25% %, 1 4/ ~20% <Box I-1 > 1

11 Box I-1 (FRB) ~25% 10~18% 4/4, 1/4 4/4 4/ / % 50% 4/ % 35%, 2000, % 50% 20% 35% 20% /4 20~25% 15~20% 2

12 , 1 2,000~2,500 S&P 500 NAPM <I-1> ) % 07% 60% -06% 11% 22% 10% 03% -13% 2) 01 05% 05% 08% -06% 03% % 06% 00% 02% 10% 02% (NAPM) : 1) )

13 % % % 08% 50% % ~8% < I-1> E (%) 2000

14 (FRB) 3 1, 10% 05~10%

15 6 12 1,, ,600

16 ,,

17 % (xpcaion) < II-1> (1985=100) : Confrnc Board 8

18 (Invnory-o- Sals raio)3/ , 32% 36% < II-2> : FRB

19 10 / / 3, ,,, NASDAQ NASDAQ ( ) (walh ffc), 10

20 11 MSDW ~3% ~10% 1/4 L,,, 1999,

21 12 M&A /4 206% -47%, (fuur mark) (FRB) IMF ~30% V U 2

22 13, (3 10 ), (Aaa Baa ), S&P500 < II-3> (Frdric S Mishkin) 1/4 53% (criical poin) 50% 50% 53% 1/4 < II-3>, 2

23 14 2 1/4 1/4 2, 1) (warning signal) 2) (fals signal) (fals signal) < II-3> / /4 1/4 53% 2 1, 3 2 2

24 % (financial accoun = ) GDP 9% % IT IT 25% IT (pric/arnings raio) (random walk) (saionary)

25 (PER) PER PER 3 < II-4>1970 S&P500 PER S&P500 PER95 99 (70) < II-4> S&P500 (PER) Gordon ( Gordon Equaion), (arning) 3 PER (nonsaionary) 16

26 (discoun ra), P = j= ρ j E + j P, ρ ( + r ), E ( )g P j + g = E 1 1 = j 1+ ρ ρ + g g E,,, (horical valu) 6% (Mhra and Prsco(1985), Wadhwani(1999), IMF(2000) ) 39%(1970 ) 17

27 40% S&P %(log ) 30% NASDAQ PER 186 PER / NASDAQ % NASDAQ 2000 PER70, S&P500 NASDAQ % 4 NADAQ S&P NASDAQ S&P500 NASDAQ S&P S&P500 NASDAQ 18

28 GDP,, (10 3 ), (Aaa - Baa ), S&P500 (Vcor Auorgrssiv Modl, VAR) VAR (Larg Macro Modl) VAR (Filardo(1999) ) (corrcion) VAR (uncondiional forcas) 30% 1/4 30% 19

29 VAR (condiional forcas),, < II-5> 2002 ( ) 2/4 3/4 < II-5> V < II-5> % 20002/ /4 8 U 20

30 21, 2/4 4/ % /4 L 90

31 Box II-1 (Rcssion Probabiliy) Confrnc Board (Composi Lading Indicaor) GDP2 (conscuiv 2-quarr dclin of GDP) NBER(Naional Burau of Economic Rsarch)BCDC(Businss Cycl Daing Commi) NBER/BCDC, 6 12,,, (rad) (broad dclin) NBER/BCDC GDP GDP ( GDP ) Nfci Modl, Lamy-Esrlla and Mishikin Modl(LEM), GDP Modl, Sock and Wason Modl(SW) (Filardo(1999) ) Probi LEM LEM (1) (0) (nonlinar) ( NBER/BCDC ) P = 1 I 2 / 2 2π, I = α + d â x ' (I -1) x Esrlla and Mishikin(1995), (10 3 ), (Moody s Baa Aaa ), S&P (n walh) MLE(Maximum Liklihood 22

32 Esimaor) β Lad Tim1 5% Probi α β 1 β 2 β 3 β 4 : /4~2000 4/4 S Louis FRB Frd DB α β ( ),,, S&P500 ( ), GDP (SW GDP ) GDP GDP (vcor auorgrssiv modl) 2 GDP GDP, 3, CPI GDP- 2001~2002 GDP GDP- GDP Esrlla- Mishikin(1995) Probi Modl GDP 23

33 24 III 1 GDP 45%(1999) 21%(1999) <III-1>1970 G-7 GDP Hodrick-Prsco Filr <III-1> G-7 G-7 053

34 <III-1> (1970~1999) : IFS 074 G7 040 <III-1> (, ) (040) G

35 < III-1> G-7 G-7 World Componn <III-1> G-7 G EU 26

36 031, 2 GDP G- 7/KR-SVAR (impuls rspons) GDP G-7/KR-SVAR G-7 GDP( ) ( ) GDP( ) (vcor auorgrssiv modl, VAR) VAR (simulanous bias) (srucural dcomposiion) III Box 3 G-7/KR-SVAR < III-2> GDP 27

37 28 1% G-7 GDP10 022~063% 3 4 G-7 GDP 67%(99), 8%(99),, 35%(99), GDP1% %GDP 3, G-7

38 < III-2> GDP (1%) 29

39 Box III-1 G-7/KR-SVAR (vcor auorgrssiv modl, VAR) Ax = C( L) x 1 + Du (II -1) x n 1 C(L)p (p-h ordr lag polynomials) n n, Dn n, u n 1 G-7 GDP ( ) 9 u D C(L) (simulanous bias) (unbiasd simaor) (consisn simaor) Brnank(1986), Sims(1986), Blanchard and Wason(1986) (srucural dcomposiion), (srucural vcor auorgrssiv modl, SVAR) (II-1) x = A C( L) x 1 1 = B ( L) x A 1 (II - 2) Du VAR u B(L)A -1 D (OLS) A -1 D (II-2) / Ó ' 1 ' ' ' 1 1 ' ' = E[ ] = A DE[ u u ] D A = A DÓ D A u 1 (II - 3) / Σ (undridnificaion problm) (,idnificaion problm) (II-3) (Cholski facorizaion) (ordring) SVAR (srucural quaion) 30

40 G-7 G-7 impac X( ), rspons Y( ) i) X X Y ii) X Y Y, Y GDP iii) X X, X x i w = α w = i γ i + β u x i i i + β w u (III- 4) w (III- 5) (III-4) GDP (x i, Sims(1980) ) u i w α i World Componn (III-5) u w GDP (wihin h currn priod) (III-5) (III-4) ( β i β w 1 ) x i α iγ j = x 1 α γ i j i i j αi + 1 α γ i i u w α γ i i u i (III- 6) (III-3) (III-1)~(III-2) 3 1/(1 α i γ i ) α i γ i < 1 1 1/(1 α i γ i ) insananous muliplir (iii) 31

41 α i γ i ( ) (iii) (III-3) α i γ j (i) (ii) γ j impac j α i rspons i j i (i)(ii) MLE(maximum liklihood simaor) α 1 α 2 α 3 α 4 α 5 α 6 α 7 α 8 β 1 γ 1 β 2 γ 2 β 3 γ 3 β 4 γ 4 β 5 γ 5 β 6 γ 6 β 7 γ 7 β 8 γ 8 β w : 1~8,,,,,,, w 2 AIC 1 2 α 5 γ 3 γ 4 (, ) (+) (β β ) Sims(1986) SVAR 32

42 1% 6% 3% 10 GDP (70~99) <III-2> : GDP 1% GDP GDP, (10 08%) World Propagaion 33

43 3 2,, GDP G-7/KR-SVAR US/KR-SVAR GDP,,,,,,, G-7/KR SVAR SVAR Box 4 US/KR-SVAR < III-3> GDP1%, GDP1% 3 08% G-7/KR 15 GDP 1% 02% %(99) GDP 1% GDP (1%) (15) (013, 99) 020%

44 , < III-3> GDP1% GDP 3 08% 3 6 GDP < III-3> GDP (1%) 35

45 < III-3> GDP (1%) 36

46 Box III-2 US/KR-SVAR G-7/KR GDP G-7/KR GDP US/KR-Macro ( ) GDP,,,,,,, 8 (SVAR) US/KR-Macro 11 ( y, i, pc, xq, imq, xp, imp, r, wq, wp, us GDP,,,,,,,,,, GDP ) y i pc xq imq xp imp r wq wp us = α 1 1 = β wq us i = δ 1 wp y = γ = κ 1 1 r u 1 xp = η = λ = µ ( xp = π = φ = φ u + α wp wp us imp r 2 i + φ u + φ pc + γ + κ 2 pc i u 2 + λ 2 imp + π 2 ) + µ (III-16) (III-17) (III-8) pc + η wp 2 wq + α wq us 3 (III- 9) r + κ + γ 3 2 wq wp ( xq wq wq + η + λ + φ + φ y wp u wq y + φ u + γ y 4 + η xp + φ 4 u wq imp (III- 7) us i xp u + φ + φ ) + φ (III-15) (III-12) imp r xq imq u u (III-13) r xq u imq (III-10) (III-11) (III-14) (III-7) GDP (III-5) (III-6) GDP (III-7) GDP (III-8) GDP (III-9) (III-10),, (III- 11) (Balassa-Samulson Thorm) 37

47 38 (III-12) GDP,, EU, (III-13) (III-14) GDP (III-4)~(III-14) MLE(Maximum Liklihood Esimaor) ( p-valu) (000) 001 (000) 003 (000) (000) (055) (000) (063) (000) 005 ) 005( ) 120( (000) (000) (041) (000) (000) (026) (005) (000) (000) (000) (000) (005) (001) (000) (000) (001) (002) (014) (000) (020) (000) (000) (000) (006) (077) (037) us us wp wp wq us wp wq r wq xq imp xp r imp wp wq r imp xp wp wq r xp imq i y r imp imq xq us wq wp xp xq pc y pc i wq i y wq pc i y u u u u u u u u u u u = = + + = + + = = = = = + = + = = 1 Sims(1986) SVAR (impuls rspons funcion)

48 39 Propagaion, < III-3> GDP 1% ( )2 23% 3 6 ( ) 2 28% 3 1%, 4 15% ( ) 2 55, ,

49 40 < III-3> 2 24% 3 08%,, (profi opporuniy) (crdi consrains) (agncy cos),

50 41 < III-3> 2 ( ) GDP 2 GNI GDP < III-3> 3 FRB 4/4 GDP 2~25% 4/4 ( )35% FRB 1~15%p,

51 42 IV 1, Humpag(2000) Humpag(2000) < IV-1>80 < IV-2>

52 < IV-1> : 1980~99, 023 < IV-2> : 1980~99 43

53 , 1980 G-7 (, ) <IV-1>80 G G <IV-1> G-7 ( )

54 45 (synchronizaion) < IV-3> 12 (moving avrag) , < IV-3>

55 IV-4,,,, 5 1% VAR- (Cholski facorizaion),,,, (liklihood raio s) 5 < IV-4> (1%),,, < IV-4> 98 (1%p) 05%p 04%p ( % ) 46

56 (duraion) Fd- Back (2000) 1% 018%, GDP014%, 065% (2000),,, GDP (vcor rror corrcion) GDP125% 0% 47

57 48,, EU -EU 1990 Saf Havn < IV-5>/ / 2001

58 < IV-5> / / (snsiiviy) (KRW/USD) = β (JPN/USD) + ε β 120 (moving avrag) (UBC) 1 / / < IV-6> / 49

59 / / + / ( KOW/USD = KOW/JPN JPN/USD ) / 100% // / / / < IV-6> / / / () / () / / 50

60 51 V % 3/4 V 20% U (sof landing) % 20~25%, 22%,

61 52, 3 6, 2001

62 < > Brnank, Bn (1989), Alrnaiv Explanaions of h Mony-Incom Corrlaion, Carngi- Rochsr Confrnc Sris on Public Policy Mhra, R and E C Prsco (1985), Th Equiy Prmium: A Puzzl, Journal of Monary Economics, Vol 15 Wadhwani, Sushil B (1999), Th US Sock Mark and h Global Economic Crisis, Naional Insiu Economic Rviw, Jan IMF(2000), World Economic Oulook, Sp Filardo, Andrw J (1999), How Rliabl Ar Rcssion Prdicion Modls?, Economics Rviw, Fdral Rsrv Bank of Kansas Ciy, 2 nd qr Esrlla, Aruro and Frdric S Mishkin (1999), Prdicing US Rcssions: Financial Variabls as Lading Indicaors, NBER Working Papr No 5379 Sims, Chrisophr A (1986), Ar Forcasing Modl Usabl for Policy Analysis?, Quarrly Rviw, Fdral Rsrv Bank of Minnapolis, Winr Blanchard, Olivr and Mark Wason (1986), Ar Businss Cycls All Alik?, Robr J Gordon, d, Th Amrican Businss Cycl Chicago: Univrsiy of Chicago Prss Humpag, Own F (2000), Forign Economic Growh and h Dollar, Economic Commnary, Fdral Rsrv Bank of Clvland, Sp 53

ๆญฏ295์œค์‚ผ์†.PDF

ๆญฏ295์œค์‚ผ์†.PDF 000 Applicaion of Coninuum Damag Modl o Lif Prdicion of Gomrically Nonlinar Srucur in High Tmpraur, 373-1,, 150 ABAQUS Chaboch Absrac In his invsigaion, h basic concp of coninuum damag mchanics was dscribd

More information

164

164 ์—๋„ˆ์ง€๊ฒฝ์ œ์—ฐ๊ตฌ์ œ 16 ๊ถŒ์ œ 1 ํ˜ธ Korean Energy Economic Review Volume 16, Number 1, March 2017 : pp. 163~190 ํ•™์ˆ  ์‹œ๋ณ€ํŒŒ๋ผ๋ฏธํ„ฐ์ผ๋ฐ˜ํ™”ํ•ด๋ฐ€ํ„ด -plucking ๋ชจํ˜•์„์ด์šฉํ•œ์ „๋ ฅ์†Œ๋น„์˜์„ ์ œ์ ๊ฒฝ๊ธฐ๊ตญ๋ฉดํŒ๋‹จํ™œ์šฉ์—ฐ๊ตฌ * 163 164 165 166 ~ 167 ln 168 [ ๊ทธ๋ฆผ 1] ์ œ์กฐ์—…์ „๋ ฅํŒ๋งค๋Ÿ‰ (a) ๋กœ๊ทธ๋ณ€ํ™˜

More information

,,,,,, KTB KOSPI 0002%, USD -0008% KTB KOSPI( ), KTB, /,, (+) KOSPI ,, KOSPI, KTB, / ( KOSPI, KTB, USD ),, 200

,,,,,, KTB KOSPI 0002%, USD -0008% KTB KOSPI( ), KTB, /,, (+) KOSPI ,, KOSPI, KTB, / ( KOSPI, KTB, USD ),, 200 CFA, (+), ETF,,,,,, ( ), (overlay) 12 1 ,,,,,, KTB KOSPI 0002%, USD -0008% KTB KOSPI( ), KTB, /,, (+) KOSPI 2000 1 4 105904,, 1999 12 16 2005 4 8 1301 KOSPI, KTB, / ( KOSPI, KTB, USD ),, 2001 44 13437

More information

(72) ๋ฐœ๋ช…์ž ์˜ค์ธํ™˜ ์„œ์šธ ๋…ธ์›๊ตฌ ์ค‘๊ณ„๋กœ 195, 101๋™ 803ํ˜ธ (์ค‘๊ณ„๋™, ์‹  ์•ˆ๋™์ง„์•„ํŒŒํŠธ) ์„œํ˜œ๋ฆฌ ์„œ์šธ ์ข…๋กœ๊ตฌ ํ‰์ฐฝ14๊ธธ 23, (ํ‰์ฐฝ๋™) ํ•œํ›ˆ์‹ ์„œ์šธ ๊ฐ•๋‚จ๊ตฌ ์–ธ์ฃผ๋กœ71๊ธธ 25-5, 301ํ˜ธ (์—ญ์‚ผ๋™, ์˜ ํ›ˆํ•˜์ด์ธ ) ์ด ๋ฐœ๋ช…์„ ์ง€์›ํ•œ ๊ตญ๊ฐ€์—ฐ๊ตฌ๊ฐœ๋ฐœ์‚ฌ์—… ๊ณผ์ œ๊ณ ์œ ๋ฒˆํ˜ธ

(72) ๋ฐœ๋ช…์ž ์˜ค์ธํ™˜ ์„œ์šธ ๋…ธ์›๊ตฌ ์ค‘๊ณ„๋กœ 195, 101๋™ 803ํ˜ธ (์ค‘๊ณ„๋™, ์‹  ์•ˆ๋™์ง„์•„ํŒŒํŠธ) ์„œํ˜œ๋ฆฌ ์„œ์šธ ์ข…๋กœ๊ตฌ ํ‰์ฐฝ14๊ธธ 23, (ํ‰์ฐฝ๋™) ํ•œํ›ˆ์‹ ์„œ์šธ ๊ฐ•๋‚จ๊ตฌ ์–ธ์ฃผ๋กœ71๊ธธ 25-5, 301ํ˜ธ (์—ญ์‚ผ๋™, ์˜ ํ›ˆํ•˜์ด์ธ ) ์ด ๋ฐœ๋ช…์„ ์ง€์›ํ•œ ๊ตญ๊ฐ€์—ฐ๊ตฌ๊ฐœ๋ฐœ์‚ฌ์—… ๊ณผ์ œ๊ณ ์œ ๋ฒˆํ˜ธ (19) ๋Œ€ํ•œ๋ฏผ๊ตญํŠนํ—ˆ์ฒญ(KR) (12) ๋“ฑ๋กํŠนํ—ˆ๊ณต๋ณด(B1) (45) ๊ณต๊ณ ์ผ์ž 2014๋…„04์›”14์ผ (11) ๋“ฑ๋ก๋ฒˆํ˜ธ 10-1384704 (24) ๋“ฑ๋ก์ผ์ž 2014๋…„04์›”07์ผ (51) ๊ตญ์ œํŠนํ—ˆ๋ถ„๋ฅ˜(Int. Cl.) F16L 9/18 (2006.01) F17D 1/00 (2006.01) F16L 3/00 (2006.01) (21) ์ถœ์›๋ฒˆํ˜ธ 10-2012-0113933

More information

ๆญฏ๊ธ€๋กœ๋ฒŒ๋””ํ”Œ๋ ˆ์ด์…˜๊ฐ€๋Šฅ์„ฑ๊ณผ์˜ํ–ฅ( ).p

ๆญฏ๊ธ€๋กœ๋ฒŒ๋””ํ”Œ๋ ˆ์ด์…˜๊ฐ€๋Šฅ์„ฑ๊ณผ์˜ํ–ฅ( ).p 23 1 V 1998,, (, %) 4 3 2 1-1 -2 211 5 9 221 5 9-2 (IMF ) -, 98 GDP, 99 ( 99, 2, 217%, 22 ) - 2 7 : Sychelles :,,,, : 22 2 199, (1993=1) 14 CRB DJ-AIG ( ) 12 DJAIG 1 CRB -74 8 6 ( ) ( : %) -23 4 199 1992

More information

ๆญฏ๋ชฉ์ฐจ.PDF

ๆญฏ๋ชฉ์ฐจ.PDF -, - 2001 5 LG i < >,, Probit, (10 T-Bond - 3 T-Bill ) 197311~19753 19801~19807 4 60%, 19817~198211 8 64% 196912~197011 19907~19913 50%, (5-1 ) ii 19922~19931 19968~19986 9 57%, 67% 19881~19897 9 41% 20007~12,

More information

Microsoft PowerPoint - F_1. 2016 ๊ฒฝ์ œ์ „๋ง_ํŽธ.pptx [์ฝ๊ธฐ ์ „์šฉ]

Microsoft PowerPoint - F_1. 2016 ๊ฒฝ์ œ์ „๋ง_ํŽธ.pptx [์ฝ๊ธฐ ์ „์šฉ] ํ—ˆ์žฌํ™˜ ()78-3 jaehwan.huh@dwsec.com ์„œ๋ฐ”์ด๋ฒŒ ๊ฒŒ์ž„ [๊ธ€๋กœ๋ฒŒ ๊ฒฝ์ œ] QE์ •์ฑ… ํ”ผ๋กœ ๋ˆ„์  ๊ณต๊ธ‰ ์ถ•์†Œ I. ์žฅ๊ธฐ ๋””ํ”Œ๋ ˆ๋Š” ์œ„ํ—˜ํ•˜๋‹ค II. ์ •์ฑ…๋ณด๋‹ค ๊ณต๊ธ‰ ์กฐ์ • III. ์ œ์กฐ์—… ๋‹ค์ด์–ดํŠธ์˜ ํšจ๊ณผ IV. ํ•œ๊ตญ๋„ ๊ฐ๋Ÿ‰์„ ์‹œ์ž‘ 8 V. 8 3 ๋‹ค์Œ ํ™˜์œจ์ „์Ÿ์˜ ์ง„์›์ง€, ์•„์‹œ์•„ [๊ฒฝ์ œ์ „๋ง/ํ™˜์œจ ์ „๋งํ‘œ] 3 [๊ธ€๋กœ๋ฒŒ ๊ฒฝ์ œ] QE์ •์ฑ… ํ”ผ๋กœ ๋ˆ„์  ๊ณต๊ธ‰ ์ถ•์†Œ ๋””ํ”Œ๋ ˆ

More information

01_ยฐรญยบร€ร‚รนรถKลก

01_ยฐรญยบร€ร‚รนรถKลก Total Return Swap 1997 7 J.P. Morgan TRSTotal Return Swap 1998 10 J.P. Morgan 75800 1996 1997 J.P. Morgan SK LG 17100 SK 16129 12593 10235 8150 7300 19981023 1 J.P. Morgan 27550 1 48250 75800 TRS SK SK

More information

ๆญฏ๋ชฉ์ฐจ.PDF

ๆญฏ๋ชฉ์ฐจ.PDF 200% 2000 9 , LG < >, IMF,, 30 IMF 30,,, 30 1997 307, 63 1999 244, 149 489% 164% 30 IMF 2000 1999 i ,, (+), (-) IMF 30 5~30 1~4 (-), (+) IMF 30, 200%, 30 ii < > < > I 30 1 1 1 2 2 3 7 4 8 5 9 II 30 11

More information

ๆญฏ

ๆญฏ (in % of EU GNI) G i n i C V u w T 1 Gini = ( 2n 2 CVuw = T = n X i i =1 n i

More information

untitled

untitled Mathematics 4 Statistics / 6. 89 Chapter 6 ( ), ( /) (Euclid geometry ( ), (( + )* /).? Archimedes,... (standard normal distriution, Gaussian distriution) X (..) (a, ). = ep{ } ฯ€ ฯƒ a 6. f ( F ( = F( f

More information

2013 ๊ฑด์„ค๊ณต์‚ฌํ‘œ์ค€ํ’ˆ์…ˆ๊ธฐ๊ณ„์„ค๋น„๋ถ€๋ฌธ ์ œ I ํŽธ ๊ณตํ†ต์‚ฌํ•ญ์ œ 1 ์žฅ ์ ์šฉ๊ธฐ์ค€์ œ 2 ์žฅ ๊ฐ€์„ค๊ณต์‚ฌ์ œ II ํŽธ ๊ธฐ๊ณ„์„ค๋น„๊ณต์‚ฌ์ œ 1 ์žฅ ๊ณตํ†ต๊ณต์‚ฌ์ œ 2 ์žฅ ๊ณต๊ธฐ์กฐํ™”์„ค๋น„๊ณต์‚ฌ์ œ 3 ์žฅ ์œ„์ƒ๋ฐ์†Œํ™”์„ค๋น„๊ณต์‚ฌ์ œ 4 ์žฅ ๊ฐ€์Šค์„ค๋น„๊ณต์‚ฌ์ œ โ…ข ํŽธ ํ”Œ๋žœํŠธ์„ค๋น„๊ณต์‚ฌ์ œ 1 ์žฅ ๊ณตํ†ต๊ณต์‚ฌ์ œ 2 ์žฅ ํ™”๋ ฅ๋ฐœ์ „๊ธฐ๊ณ„์„ค๋น„๊ณต์‚ฌ์ œ 3 ์žฅ ์ˆ˜๋ ฅ๋ฐœ์ „๊ธฐ๊ณ„์„ค๋น„๊ณต์‚ฌ์ œ 4 ์žฅ ์ œ์ฒ ๊ธฐ๊ณ„์„ค๋น„๊ณต์‚ฌ์ œ 5 ์žฅ ์“ฐ๋ ˆ๊ธฐ์†Œ๊ฐ๊ธฐ๊ณ„์„ค๋น„๊ณต์‚ฌ์ œ

More information

์ค‘์†Œ๊ธฐ์—…๊ฒฝ๊ธฐ์ง€์ˆ˜๋ฐ๊ฒฝ์˜ํ™˜๊ฒฝ์ง€์ˆ˜ ๊ฐœ๋ฐœ์—๊ด€ํ•œ์—ฐ๊ตฌ - ์ œ์กฐ์—…์ค‘์‹ฌ์œผ๋กœ - A Study on Development of the Business Indicators in SMEs focused on manufacturing ์š”์•ฝ 1) 125 IPISA 124 ISISA 120 120 115 110 105 100 95 116 112 108 104

More information

<4D6963726F736F667420576F7264202D20C1A4B3E2BFACC0E5BFA120B4EBBAF1C7D120C0CEB7C220BFEEBFB520C0FCB7AB5F20504446>

<4D6963726F736F667420576F7264202D20C1A4B3E2BFACC0E5BFA120B4EBBAF1C7D120C0CEB7C220BFEEBFB520C0FCB7AB5F20504446> ์ •๋…„ ์—ฐ์žฅ์— ๋Œ€๋น„ํ•œ ์ธ๋ ฅ ์šด์˜ ์ „๋žต - ์ž„๊ธˆ ํ”ผํฌ์ œ๋ฅผ ํ†ตํ•œ ์ ‘๊ทผ ํƒ€์›Œ์Šค ์™“์Šจ ์ตœํ˜„์•„ ๋ถ€์‚ฌ์žฅ ๊ตญ๋‚ด ๊ธ‰์—ฌ์ฒด๊ณ„์˜ ํŠน์ง•: ์—ฐ๊ณต์„œ์—ด์  ๊ธ‰์—ฌ์ฒด๊ณ„ IMF ์ดํ›„ ์—ฐ๋ด‰์ œ ๋„์ž…, ์ง๊ธ‰๋ณ„ ๊ธ‰์—ฌ์˜ ์ƒํ•˜ํ•œ์„ ์ •ํ•˜๋Š” ๋“ฑ ์—ฌ๋Ÿฌ ๊ฐ€์ง€ ์ƒˆ๋กœ์šด ์ œ๋„๋“ค๋กœ ์ธํ•ด ์šฐ๋ฆฌ๋‚˜๋ผ์˜ ๊ธ‰์—ฌ์ฒด๊ณ„๊ฐ€ ๊ณผ๊ฑฐ์— ๋น„ํ•ด ์—ฐ๊ณต์„œ์—ด์ ์ธ ์š”์†Œ์—์„œ ๋งŽ์ด ํƒˆํ”ผํ–ˆ๋‹ค๊ณ  ํ•˜๋‚˜ ์ƒ์‚ฐ์ง์€ ์—ฌ์ „ํžˆ ์—ฐ๊ณต์„œ์—ด์ ์ธ ํ˜ธ๋ด‰์ œ๊ฐ€ ์‚ฌ์šฉ๋˜๊ณ  ์žˆ๊ณ , ์‚ฌ๋ฌด์ง

More information

untitled

untitled 4 (YOY) (%) 3 1% 3.5 3 2 8% 8 5 1 29 9 16 12 16 2.5 1 6% 2 1.5-1 4% 1.5 CPI() - () -1 1-1 2-1 3-1 4-1 5-1 6-1 7-1 8-1 -2-3 2% % 5-1 5-1 5-19 5-28 6-6 6-15 6-24 8 6 4 2-2 -4-6 3-2 3-1 4-6 5-2 5-1 6-6 7-2

More information

ๆญฏ๊ฒฝ์ œ.PDF

ๆญฏ๊ฒฝ์ œ.PDF LG 2% 8 6 ( : %) 22 23 149 136 4 2 37 19 68 68 33 76-2 21 22 23 : -4-7 4-7 2 1/3 3% 14 12 ( ) 15 1 4 2 1 98 5 96 94 ( ) 2 21 22 23-5 -2-4 1998 1999 2 21 22 23 2/3 (1) 21, 11, 1, ( ) 1 9 65 6 55 9, 5 8,

More information

์‹ ํฅ๊ตญ ์†Œ๋น„ํ™•๋Œ€๋กœ ๊ธ€๋กœ๋ฒŒ ๋ฌด์—ญ ๋ถˆ๊ท ํ˜•์ด ๊ตฌ์กฐ์ ์œผ๋กœ ์™„ํ™”๋œ๋‹ค. ๊ธ€๋กœ๋ฒŒ ๊ฒฝ๊ธฐ ํšŒ๋ณต๋˜๋Š” ๊ฐ€์šด๋ฐ ์•„์‹œ์•„ ์‹ ํฅ๊ตญ ๊ฒฌ์กฐํ•œ ์„ฑ์žฅ ์˜ˆ์ƒ ๊ธ€๋กœ๋ฒŒ ๊ฒฝ์ œ๋Š” ๊ธˆ์œต์œ„๊ธฐ ์ด์ „์— ๋น„ํ•ด์„œ๋Š” ๊ฒฝ์ œ์„ฑ์žฅ๋ฅ ์ด ํ•˜๋ฝํ•œ ์ €์„ฑ์žฅ ๊ณผ์ •์ด ์˜ˆ์ƒ๋˜์ง€๋งŒ ๋…„์„ ์ €์ ์„ ํ†ต๊ณผํ•ด ์˜ฌํ•ด๋ถ€ํ„ฐ๋Š” ํšŒ๋ณต์„ธ๊ฐ€ ์ ์ฐจ ์•ˆ์ •ํ™”๋  ์ „๋ง์ด๋‹ค.

์‹ ํฅ๊ตญ ์†Œ๋น„ํ™•๋Œ€๋กœ ๊ธ€๋กœ๋ฒŒ ๋ฌด์—ญ ๋ถˆ๊ท ํ˜•์ด ๊ตฌ์กฐ์ ์œผ๋กœ ์™„ํ™”๋œ๋‹ค. ๊ธ€๋กœ๋ฒŒ ๊ฒฝ๊ธฐ ํšŒ๋ณต๋˜๋Š” ๊ฐ€์šด๋ฐ ์•„์‹œ์•„ ์‹ ํฅ๊ตญ ๊ฒฌ์กฐํ•œ ์„ฑ์žฅ ์˜ˆ์ƒ ๊ธ€๋กœ๋ฒŒ ๊ฒฝ์ œ๋Š” ๊ธˆ์œต์œ„๊ธฐ ์ด์ „์— ๋น„ํ•ด์„œ๋Š” ๊ฒฝ์ œ์„ฑ์žฅ๋ฅ ์ด ํ•˜๋ฝํ•œ ์ €์„ฑ์žฅ ๊ณผ์ •์ด ์˜ˆ์ƒ๋˜์ง€๋งŒ ๋…„์„ ์ €์ ์„ ํ†ต๊ณผํ•ด ์˜ฌํ•ด๋ถ€ํ„ฐ๋Š” ํšŒ๋ณต์„ธ๊ฐ€ ์ ์ฐจ ์•ˆ์ •ํ™”๋  ์ „๋ง์ด๋‹ค. Note.. Economist ์œ ์ต์„  )-, sun.yoo@wooriwm.com ์•ˆ๊ธฐํƒœ )-, kt.an@wooriwm.com ์œ ์ต์„ ์˜ ๊ฒฝ์ œ๋ณ„๊ณก( ็ถ“ ๆฟŸ ๅˆฅ ๆ›ฒ ): Imbalance์˜ Balance๊ฐ€ Great Rotation์— ๋ฏธ์น˜๋Š” ์˜ํ–ฅ ๊ธ€๋กœ๋ฒŒ ๋ถˆ๊ท ํ˜• ์™„ํ™”๊ฐ€ ๊ธ€๋กœ๋ฒŒ ์ž๊ธˆํ๋ฆ„์˜ ๋ณ€ํ™”๋ฅผ ์ด‰๋ฐœ์‹œํ‚ฌ ์ „๋ง ํ–ฅํ›„ ์‹ ํฅ๊ตญ ๋‚ด์ˆ˜ ํ™•๋Œ€๊ฐ€ ๋ณธ๊ฒฉํ™”๋˜๋ฉด์„œ ๊ธ€๋กœ๋ฒŒ ๋ฌด์—ญ ๋ถˆ๊ท ํ˜•์ด

More information

100, Jan. 21, ํ˜ธ, Jan. 21, , Jan. 21, 2005

100, Jan. 21, ํ˜ธ, Jan. 21, , Jan. 21, 2005 100 Bond Issue Performance Evaluation Risk management 100, Jan. 21, 2005 100, Jan. 21, 200 5 100ํ˜ธ, Jan. 21, 2005 2 100, Jan. 21, 2005 3 100, Jan. 21, 2005 4 100, Jan. 21, 2005 5 100, Jan. 21, 2005 6 100,

More information

๋ณธ ๋ฐœ๋ช…์€ ์ค‘๊ณต์ฝ”์–ด ํ”„๋ฆฌ์บ์ŠคํŠธ ์Šฌ๋ž˜๋ธŒ ๋ฐ ๊ทธ ์‹œ๊ณต๋ฐฉ๋ฒ•์— ๊ด€ํ•œ ๊ฒƒ์œผ๋กœ, ์ž์„ธํ•˜๊ฒŒ๋Š” ์ค‘๊ณต์ฝ”์–ด๋กœ ํ˜•์„ฑ๋œ ํ”„๋ฆฌ์บ์ŠคํŠธ ์Šฌ๋ž˜๋ธŒ ์— ์˜จ๋Œ์„ ์ผ์ฒด๋กœ ๊ตฌ์„ฑํ•œ ์Šฌ๋ž˜๋ธŒ ๊ตฌ์กฐ ๋ฐ ๊ทธ ์‹œ๊ณต๋ฐฉ๋ฒ•์— ๊ด€ํ•œ ๊ฒƒ์ด๋‹ค. ์ด๋ฅผ ์œ„ํ•œ ์˜จ๋Œ ์ผ์ฒดํ˜• ์ค‘๊ณต์ฝ”์–ด ํ”„๋ฆฌ์บ์ŠคํŠธ ์Šฌ๋ž˜๋ธŒ๋Š”, ๊ณต์žฅ์—์„œ ์ œ์ž‘๋˜๋Š” ์ค‘๊ณต์ฝ”์–ด ํ”„

๋ณธ ๋ฐœ๋ช…์€ ์ค‘๊ณต์ฝ”์–ด ํ”„๋ฆฌ์บ์ŠคํŠธ ์Šฌ๋ž˜๋ธŒ ๋ฐ ๊ทธ ์‹œ๊ณต๋ฐฉ๋ฒ•์— ๊ด€ํ•œ ๊ฒƒ์œผ๋กœ, ์ž์„ธํ•˜๊ฒŒ๋Š” ์ค‘๊ณต์ฝ”์–ด๋กœ ํ˜•์„ฑ๋œ ํ”„๋ฆฌ์บ์ŠคํŠธ ์Šฌ๋ž˜๋ธŒ ์— ์˜จ๋Œ์„ ์ผ์ฒด๋กœ ๊ตฌ์„ฑํ•œ ์Šฌ๋ž˜๋ธŒ ๊ตฌ์กฐ ๋ฐ ๊ทธ ์‹œ๊ณต๋ฐฉ๋ฒ•์— ๊ด€ํ•œ ๊ฒƒ์ด๋‹ค. ์ด๋ฅผ ์œ„ํ•œ ์˜จ๋Œ ์ผ์ฒดํ˜• ์ค‘๊ณต์ฝ”์–ด ํ”„๋ฆฌ์บ์ŠคํŠธ ์Šฌ๋ž˜๋ธŒ๋Š”, ๊ณต์žฅ์—์„œ ์ œ์ž‘๋˜๋Š” ์ค‘๊ณต์ฝ”์–ด ํ”„ (51) Int. Cl. E04B 5/32 (2006.01) (19)๋Œ€ํ•œ๋ฏผ๊ตญํŠนํ—ˆ์ฒญ(KR) (12) ๋“ฑ๋กํŠนํ—ˆ๊ณต๋ณด(B1) (45) ๊ณต๊ณ ์ผ์ž (11) ๋“ฑ๋ก๋ฒˆํ˜ธ (24) ๋“ฑ๋ก์ผ์ž 2007๋…„03์›”12์ผ 10-0693122 2007๋…„03์›”05์ผ (21) ์ถœ์›๋ฒˆํ˜ธ 10-2006-0048965 (65) ๊ณต๊ฐœ๋ฒˆํ˜ธ (22) ์ถœ์›์ผ์ž 2006๋…„05์›”30์ผ (43) ๊ณต๊ฐœ์ผ์ž ์‹ฌ์‚ฌ์ฒญ๊ตฌ์ผ์ž

More information

i f i f (disposition effect) ์˜ํ™•๋ฅ  ์˜ํ™•๋ฅ  i f i f i f i f i f i f GARCH-in-Mean GARCH-in-Mean , ( ) ( ). ( ), / ( ), (1 ), S&P500,,. ์ƒ์Šน๋ฐ˜์‘๊ณ„์ˆ˜ ๋กœ ~2008.12 15) ~ ~ m 10 20 (8) (10) (11) (8) (10) (11) 0.011 (0.23)

More information

1 n dn dt = f v = 4 ฯ€ m 2kT 3/ 2 v 2 mv exp 2kT 2 f v dfv = 0 v = 0, v = /// fv = max = 0 dv 2kT v p = m 1/ 2 vfvdv 0 2 2kT = = vav = v f dv ฯ€ m

1 n dn dt = f v = 4 ฯ€ m 2kT 3/ 2 v 2 mv exp 2kT 2 f v dfv = 0 v = 0, v = /// fv = max = 0 dv 2kT v p = m 1/ 2 vfvdv 0 2 2kT = = vav = v f dv ฯ€ m n dn dt f v 4 ฯ€ m kt 3/ v mv exp kt f v dfv 0 v 0, v /// fv max 0 dv kt v p m / vfvdv 0 kt vav. 8v f dv ฯ€ m k m 0 v / R0 4 T vav.45 0 cm / sec M M p v v fvdv 0 3 fvdv 0 kt m / 3kT v v. 5 m rms v p n dn

More information

์ œ24์กฐ

์ œ24์กฐ ๊ณ ์šฉ๋ฅ  ์ œ๊ณ ๋ฅผ ์œ„ํ•œ ๋…ธ๋™์‹œ์žฅ ๊ฐœ์„  ๋งค๋‰ด์–ผ ์ œ 2 ๊ถŒ ๋ถ€ ๋ก ๋ณ€์–‘๊ทœ ์ด์Šน๊ธธ ๋‚จ์žฌ๋Ÿ‰ ์™ธ ์ œ1์žฅ ๋„์ž… ๋ฐ ์š”์•ฝ ์ œ2์žฅ ๊ณ ์šฉ์ฐฝ์ถœ๋ ฅ ๊ฐœ์„ ๋ฐฉ์•ˆ ์ œ3์žฅ ๋…ธ๋™์‹œ์žฅ์ œ๋„ ๊ฐœ์„ ๋ฐฉ์•ˆ ๊ณ ์šฉ๋ฅ  ์ œ๊ณ ๋ฅผ ์œ„ํ•œ ๋…ธ๋™์‹œ์žฅ ๊ฐœ์„  ๋งค๋‰ด์–ผ ์ œ2๊ถŒ ๋ถ€ ๋ก 1ํŒ1์‡„ ์ธ์‡„ / 2011๋…„ 1์›” 18์ผ 1ํŒ1์‡„ ๋ฐœํ–‰ / 2011๋…„ 1์›” 21์ผ ๋ฐœํ–‰์ฒ˜ ํ•œ๊ตญ๊ฒฝ์ œ์—ฐ๊ตฌ์› ๋ฐœํ–‰์ธ ๊น€์˜์šฉ ํŽธ์ง‘์ธ ๊น€์˜์šฉ ๋“ฑ๋ก๋ฒˆํ˜ธ ์ œ318

More information

แ„’แ…กแ„‡แ…กแ†ซแ„€แ…ต_แ„‘แ…ญแ„Œแ…ต

แ„’แ…กแ„‡แ…กแ†ซแ„€แ…ต_แ„‘แ…ญแ„Œแ…ต LEG WORKING PAPER SERIES 2012_ 05 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 ร รถ 37 38 39 40 41 42 43 44 45 Discussion Paper 49 50 51 LEG WORKING PAPER

More information

Microsoft PowerPoint - ํ™•๋ฅ 3์žฅ-1-v2007 [ํ˜ธํ™˜ ๋ชจ๋“œ]

Microsoft PowerPoint - ํ™•๋ฅ 3์žฅ-1-v2007 [ํ˜ธํ™˜ ๋ชจ๋“œ] Ch.3 Random Variables ๋ถˆ๊ทœ์น™๋ณ€์ˆ˜ 3. The Noion of a Random Variable S ฮถ ฮถ real line Random variable a funcion ha assigns a real number ฮถ o each oucome ฮถ in he sample space of a random eperimen Specificaion of

More information

<30392D323020B1C7B1E2BCF628C7D1B1B9B1E2BEF7C0C720B4EBC1DFB3B2B9CC292E687770>

<30392D323020B1C7B1E2BCF628C7D1B1B9B1E2BEF7C0C720B4EBC1DFB3B2B9CC292E687770> ์—ฐ๊ตฌ๋ณด๊ณ ์„œ 09-20 ํ•œ๊ตญ๊ธฐ์—…์˜ ๋Œ€์ค‘๋‚จ๋ฏธ ํˆฌ์ž์ง„์ถœ ์„ฑ๊ณผ์™€ ๊ณผ์ œ ๊ถŒ๊ธฐ์ˆ˜ ๊น€์ง„์˜ค ๊ณ ํฌ์ฑ„ ํ•œ๊ตญ๊ธฐ์—…์˜ ๋Œ€์ค‘๋‚จ๋ฏธ ํˆฌ์ž์ง„์ถœ ์„ฑ๊ณผ์™€ ๊ณผ์ œ ์—ฐ๊ตฌ๋ณด๊ณ ์„œ 09-20 ํ•œ๊ตญ๊ธฐ์—…์˜ ๋Œ€์ค‘๋‚จ๋ฏธ ํˆฌ์ž์ง„์ถœ ์„ฑ๊ณผ์™€ ๊ณผ์ œ ๊ถŒ๊ธฐ์ˆ˜ใ†๊น€์ง„์˜คใ†๊ณ ํฌ์ฑ„ ์„œ ์–ธ ์ „ํ†ต์ ์œผ๋กœ ์šฐ๋ฆฌ์—๊ฒŒ ์ค‘๋‚จ๋ฏธ์‹œ์žฅ์€ ์ˆ˜์ถœ๊ทœ๋ชจ๋Š” ํฌ์ง€ ์•Š์ง€๋งŒ ๋ฌด์—ญํ‘์ž์‹œ์žฅ์œผ๋กœ ์ธ ์‹๋˜์–ด ์™”์Šต๋‹ˆ๋‹ค. ํŠนํžˆ 2000๋…„๋Œ€ ๋“ค์–ด ๋Œ€์ค‘๋‚จ๋ฏธ ์ˆ˜์ถœ์ด ํฐ ํญ์˜ ์„ฑ์žฅ์„ธ๋ฅผ

More information

' ' ( ) 2 60 ' ' ( ) 50% 21% 95 ( ) 43% 13% 10 5 ( ) 24% 9% 11 5 ( ) 20%6% ' ' ' ' ' ' 1965

' ' ( ) 2 60 ' ' ( ) 50% 21% 95 ( ) 43% 13% 10 5 ( ) 24% 9% 11 5 ( ) 20%6% ' ' ' ' ' ' 1965 I 10 20005 ' ' II 1 1960 1 1) 19659 ' ' ( ) 2 60 ' ' 60 75 85 (1965-1970) 50% 21% 95 (1971-1975) 43% 13% 10 5 (1976-1980) 24% 9% 11 5 (1981-1985) 20%6% 1960 1970 1970 3 ' ' ' ' ' ' 1965 ' ' 'A N ' (NEP)

More information

Microsoft Word - 09๋…„ 2์›” ๊ฒฝ์ œ๋™ํ–ฅ.doc

Microsoft Word - 09๋…„ 2์›” ๊ฒฝ์ œ๋™ํ–ฅ.doc ๊ฒฝ์ œ, ์ฆ์‹œ๋ถ„์„ ๋ฐ ์ „๋ง ๋ฏธ๊ตญ ๊ฒฝ์ œํ˜„ํ™ฉ๊ณผ ์ „๋ง Insightful & easy-to-understandable market outlook for the serious investors ์‹ค๋ฌผ๊ฒฝ๊ธฐ ์นจ์ฒด ์‹ฌํ™”, ๊ทธ ํญ๊ณผ ๊นŠ์ด๋Š”? I. ๊ฒฝ์ œ๋™ํ–ฅ ์‹ค๋ฌผ๊ฒฝ๊ธฐ ์นจ์ฒด ๋ณธ๊ฒฉํ™” - 2008๋…„ 4์‚ฌ๋ถ„๊ธฐ ๋ฏธ๊ตญ์˜ ์‹ค์งˆ GDP๊ฒฝ์ œ์„ฑ์žฅ๋ฅ ์€ ์ „๋…„๋Œ€๋น„ -3.8%์˜ ๋งˆ์ด๋„ˆ์Šค ์„ฑ์žฅ์„ธ๋ฅผ ๋ณด์ด๋ฉฐ

More information

01011008-1003200800a.ps

01011008-1003200800a.ps 4 2008๋…„ 3์›” 10์ผ ์›”์š”์ผ ์ƒˆ ์ •๋ถ€ ์ถœ๋ฒ”, ๊ต์œก ์ •์ฑ… ์–ด๋–ป๊ฒŒ ๋‹ฌ๋ผ์ง€๋‚˜ ์‹ ์ž„ ๊ต์œก๊ฐ์—๊ฒŒ ๋“ฃ๋Š”๋‹ค - ๋Œ€์ž… ์ž์œจํ™” ์˜์–ด ๊ณต๊ต์œก ๊ฐ•ํ™” ์ดˆ์  ์ œ์ฃผ ์ž์œจ์  ์—ญ๋Ÿ‰ ๋ฐœํœ˜ํ•  ์ˆ˜ ์žˆ๊ฒŒ ๊ต์œก์—…๋ฌด ํ•™๊ต๋กœ ๋Œ€ํญ ์ด์–‘ ์ƒˆ๋กœ์šด ์ •๋ถ€์˜ ์ถœ๋ฒ”๊ณผ ํ•จ๊ป˜ ๊ต์œก์ •์ฑ…์—๋„ ์ผ๋Œ€ ๋ณ€ํ™”๊ฐ€ ์˜ˆ๊ณ ๋˜ ๊ณ  ์žˆ๋‹ค. ์ง€๋‚œ๋‹ฌ 25์ผ ์ทจ์ž„์‹์—์„œ ์ด๋ช…๋ฐ• ๋Œ€ํ†ต๋ น์€ ๊ต์œก๊ฐœํ˜์€ ๋ฌด์—‡๋ณด๋‹ค ์‹œ๊ธ‰ํ•˜๋‹ค ๋ฉฐ ๊ธ€๋กœ๋ฒŒ ์Šคํƒ ๋”๋“œ๋ฅผ

More information

์ด์Šˆ๋ถ„์„ 2000 Vol.1

์ด์Šˆ๋ถ„์„ 2000 Vol.1 i ii iii iv 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

More information

๊ฐ€๋ณ๊ฒŒ์ฝ๋Š”-๋‚ด์ง€-1-2

๊ฐ€๋ณ๊ฒŒ์ฝ๋Š”-๋‚ด์ง€-1-2 I 01. 10 11 12 02. 13 14 15 03. 16 17 18 04. 19 20 21 05. 22 23 24 06. 25 26 27 07. 28 29 08. 30 31 09. 32 33 10. 34 35 36 11. 37 12. 38 13. 39 14. 40 15. 41 16. 42 43 17. 44 45 18. 46 19. 47 48 20. 49

More information

kbs_thesis.hwp

kbs_thesis.hwp - I - - II - - III - - IV - - 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 -

More information

ํ•œ๋ˆˆ์—-์•„์„ธ์•ˆ ๋‚ด์ง€-1

ํ•œ๋ˆˆ์—-์•„์„ธ์•ˆ ๋‚ด์ง€-1 I 12 I 13 14 I 15 16 I 17 18 II 20 II 21 22 II 23 24 II 25 26 II 27 28 II 29 30 II 31 32 II 33 34 II 35 36 III 38 III 39 40 III 41 42 III 43 44 III 45 46 III 47 48 III 49 50 IV 52 IV 53 54 IV 55 56 IV

More information

15์ธํ”Œ๋ ˆ์ด์…˜01-๋ชฉ์ฐจ1~9

15์ธํ”Œ๋ ˆ์ด์…˜01-๋ชฉ์ฐจ1~9 ISSN 87-381 15. 1 15. 1 13 1 1.3 1. 1.8 1.5 1. 1.1 () 1.5 1..1 1.8 1.7 1.3 () 1..7.6...3 (). 1.5 3.6 3.3.9. 6.3 5.5 5.5 5.3.9.9 ().6.3.. 1.6 1. i 6 5 6 5 5 5 3 3 3 3 1 1 1 1-1 -1 13 1 1).6..3.1.3.

More information

2/21

2/21 ์ง€์ฃผํšŒ์‚ฌ LG์˜ ์„ค๋ฆฝ๊ณผ์ • ๋ฐ ํŠน์ง• ์†Œ์œ ๊ตฌ์กฐ๋ฅผ ์ค‘์‹ฌ์œผ๋กœ ์ด์€์ •_์ข‹์€๊ธฐ์—…์ง€๋ฐฐ๊ตฌ์กฐ์—ฐ๊ตฌ์†Œ ๊ธฐ์—…์ •๋ณด์‹ค์žฅ ์ด์ฃผ์˜_์ข‹์€๊ธฐ์—…์ง€๋ฐฐ๊ตฌ์กฐ์—ฐ๊ตฌ์†Œ ์—ฐ๊ตฌ์› 1/21 2/21 3/21 4/21 5/21 6/21 7/21 8/21 9/21 10/21 11/21 12/21 13/21 14/21 15/21 16/21 17/21 18/21 19/21 20/21 [๋ณ„์ฒจ1] 2000.12.31.ํ˜„์žฌ

More information

์„œ๋น„์Šค ์ˆ˜์ถœ์— ์žˆ์–ด์„œ๋„ ์ผ๋ณธ์˜ ํ•ด์™ธ์ง„์ถœ์ด ๋’ค์ฒ˜์ง€๊ณ  ์žˆ๋‹ค๊ณ  ์ง€์  (IMF ์—ฐ๋ก€ ์ผ๋ณธ๊ฒฝ์ œ ๋ณด๊ณ ์„œ ์ฃผ์š”๋‚ด์šฉ) IMF ์—ฐ๋ก€ ์ผ๋ณธ๊ฒฝ์ œ ๋ณด๊ณ ์„œ์—์„œ ๆ—ฅ ้Š€ ์˜ ๋Œ€ ๊ทœ๋ชจ ๊ธˆ์œต์™„ํ™”๋ฅผ ํ†ตํ•ด ๋ฌผ๊ฐ€๊ฐ€ (์ „๋…„๋Œ€๋น„) ์ƒ์Šนํ–ˆ๋‹ค๊ณ  ๋ถ„์„ํ•˜์˜€์œผ๋ฉฐ, ์ผ๋ณธ์˜ ์†Œ๋น„์„ธ ์ธ์ƒ ์žฌ์—ฐ๊ธฐ(2017.4์›” 2019.1

์„œ๋น„์Šค ์ˆ˜์ถœ์— ์žˆ์–ด์„œ๋„ ์ผ๋ณธ์˜ ํ•ด์™ธ์ง„์ถœ์ด ๋’ค์ฒ˜์ง€๊ณ  ์žˆ๋‹ค๊ณ  ์ง€์  (IMF ์—ฐ๋ก€ ์ผ๋ณธ๊ฒฝ์ œ ๋ณด๊ณ ์„œ ์ฃผ์š”๋‚ด์šฉ) IMF ์—ฐ๋ก€ ์ผ๋ณธ๊ฒฝ์ œ ๋ณด๊ณ ์„œ์—์„œ ๆ—ฅ ้Š€ ์˜ ๋Œ€ ๊ทœ๋ชจ ๊ธˆ์œต์™„ํ™”๋ฅผ ํ†ตํ•ด ๋ฌผ๊ฐ€๊ฐ€ (์ „๋…„๋Œ€๋น„) ์ƒ์Šนํ–ˆ๋‹ค๊ณ  ๋ถ„์„ํ•˜์˜€์œผ๋ฉฐ, ์ผ๋ณธ์˜ ์†Œ๋น„์„ธ ์ธ์ƒ ์žฌ์—ฐ๊ธฐ(2017.4์›” 2019.1 ์ผ๋ณธ๊ฒฝ์ œ ์ฃผ๊ฐ„๋™ํ–ฅ(6.18-24) ํ•ต์‹ฌ๋‚ด์šฉ 1. ๊ฒฝ์ œ์ผ๋ฐ˜ (์˜๊ตญ์˜ EU ์ž”๋ฅ˜์‹œ์—๋„ ์—”ํ™” ๊ฐ•์„ธ ๊ฐ€๋Šฅ์„ฑ) ์˜๊ตญ์˜ EU ํƒˆํ‡ด ์—ฌ๋ถ€์— ๋Œ€ํ•œ ์—ฌ๋ก ์กฐ์‚ฌ ๋“ฑ์—์„œ EU ์ž”๋ฅ˜ ์ง€์ง€๊ฐ€ ์šฐ์„ธ๋ฅผ ๋ณด์ด๊ณ  ์žˆ์Œ์—๋„ ๋ถˆ๊ตฌ, ์—”ํ™”๋งค๋„ ๋‹ฌ๋Ÿฌ๋งค์ž… ์›€์ง์ž„ ํ™•์‚ฐ์€ ์ œํ•œ์ ์œผ๋กœ, ์—”ํ™”๊ฐ•์„ธ ์ „๋ง์ด ์ง€์† ใ…‡ ๊ตญ๋ฏผํˆฌํ‘œ๋ฅผ ํ†ตํ•ด ์˜๊ตญ์˜ EU ์ž”๋ฅ˜๊ฐ€ ๊ฒฐ์ •๋˜์–ด๋„ ๋Œ€ํญ์ ์ธ ์—”ํ™”์•ฝ์„ธ๋Š” ๊ธฐ๋Œ€ํ•˜๊ธฐ ์–ด๋ ค์šธ ๊ฒƒ(๊ณ ๋ฐ”์•ผ์‹œ

More information

โ… . ๊ธˆ์œต์œ„๊ธฐ ๋ฒ—์–ด๋‚˜๋‚˜ 28๋…„ 1์›”๊ณผ 29๋…„ 2์›” ๋‘ ์ฐจ๋ก€์— ๊ฑธ์ณ ์ค‘๋Œ€ํ•œ ์œ„๊ธฐ ๊ตญ๋ฉด์„ ๋งž์•˜๋˜ ๊ธ€๋กœ๋ฒŒ ๊ธˆ์œต ์‹œ์žฅ์ด 3์›” ์ดํ›„ ๋น ๋ฅด๊ฒŒ ์•ˆ์ •์„ ๋˜์ฐพ์•„๊ฐ€๊ณ  ์žˆ๋‹ค. ๋ฏธ๊ตญ์„ ๋น„๋กฏํ•œ ๊ฐ๊ตญ ์ •๋ถ€์™€ ์ค‘์•™ ์€ํ–‰์ด ์œ ๋™์„ฑ ๊ณต๊ธ‰์„ ํฌ๊ฒŒ ๋Š˜๋ฆฌ๊ณ  ๊ธˆ์œต๊ธฐ๊ด€์˜ ์ž๋ณธํ™•์ถฉ์„ ์ง€์›ํ•˜๋ฉด์„œ ๊ธˆ์œตํšŒ์‚ฌ๋“ค ์˜

โ… . ๊ธˆ์œต์œ„๊ธฐ ๋ฒ—์–ด๋‚˜๋‚˜ 28๋…„ 1์›”๊ณผ 29๋…„ 2์›” ๋‘ ์ฐจ๋ก€์— ๊ฑธ์ณ ์ค‘๋Œ€ํ•œ ์œ„๊ธฐ ๊ตญ๋ฉด์„ ๋งž์•˜๋˜ ๊ธ€๋กœ๋ฒŒ ๊ธˆ์œต ์‹œ์žฅ์ด 3์›” ์ดํ›„ ๋น ๋ฅด๊ฒŒ ์•ˆ์ •์„ ๋˜์ฐพ์•„๊ฐ€๊ณ  ์žˆ๋‹ค. ๋ฏธ๊ตญ์„ ๋น„๋กฏํ•œ ๊ฐ๊ตญ ์ •๋ถ€์™€ ์ค‘์•™ ์€ํ–‰์ด ์œ ๋™์„ฑ ๊ณต๊ธ‰์„ ํฌ๊ฒŒ ๋Š˜๋ฆฌ๊ณ  ๊ธˆ์œต๊ธฐ๊ด€์˜ ์ž๋ณธํ™•์ถฉ์„ ์ง€์›ํ•˜๋ฉด์„œ ๊ธˆ์œตํšŒ์‚ฌ๋“ค ์˜ LGERI ๋ฆฌํฌํŠธ 29๋…„ ํ•˜๋ฐ˜๊ธฐ ๊ตญ๋‚ด์™ธ ๊ฒฝ์ œ์ „๋ง ๊ฒฝ์ œ์—ฐ๊ตฌ์‹ค gtlee@lgeri.com ๊ธˆ์œต์—ฐ๊ตฌ์‹ค hybae@lgeri.com โ… . ๊ธˆ์œต์œ„๊ธฐ ๋ฒ—์–ด๋‚˜๋‚˜ โ…ก. ์„ธ๊ณ„๊ฒฝ์ œ์˜ ํ–ฅ๋ฐฉ์€ โ…ข. ๊ตญ๋‚ด๊ฒฝ๊ธฐ ์–ธ์ œ ํšŒ๋ณต๋˜๋‚˜ ๊ธˆ์œต๊ธฐ๊ด€ ์†์‹ค์ด ์ค„์–ด๋“ค๊ณ  ์œ„ํ—˜ํšŒํ”ผ ์„ฑํ–ฅ์ด ์™„ํ™”๋˜๋ฉด์„œ ๊ธˆ์œต์‹œ์žฅ์ด ๊ทน์‹ฌํ•œ ๋ถˆ์•ˆ์œผ๋กœ๋ถ€ํ„ฐ ๋ฒ—์–ด ๋‚˜๊ณ  ์žˆ๋‹ค. ๋ฏธ๊ตญ ์ฃผํƒ๊ฐ€๊ฒฉ์˜ ํ•˜๋ฝ์†๋„๊ฐ€ ๋‘”ํ™”๋˜๋Š” ์กฐ์ง๋„ ๋‚˜ํƒ€๋‚˜๊ณ  ์žˆ์ง€๋งŒ,

More information

Microsoft Word - ๋™ํƒœ์  ๋ชจํ˜•.doc

Microsoft Word - ๋™ํƒœ์  ๋ชจํ˜•.doc ๋™ํƒœ์ ๋ชจํ˜• - ์‹œ์ฐจ๋ถ„ํฌ๋ชจํ˜• (lag disribued model) I. ๊ฐœ์š” A. ๊ฒฝ์ œ์ ํ–‰์œ„๋‚˜๊ฒฐ์ •๋“ค์˜ํšจ๊ณผ๋Š”์ฆ‰์‹œ์ ์œผ๋กœ๋‹ค๋‚˜ํƒ€๋‚˜์ง€์•Š๊ณ ๋ฏธ๋ž˜์˜์ƒ๋‹น๊ธฐ๊ฐ„๋™์•ˆ๋ถ„ํฌ๋จ i. ๊ธฐ์˜ํ–‰์œ„๋‚˜๊ฒฐ์ •๋“ค์ด ๊ธฐ๋ฟ์•„๋‹ˆ๋ผ + ๊ธฐ, + ๊ธฐ๋“ฑ์—๋„์˜ํ–ฅ์„๋ฏธ์น˜๋Š”๊ฒฝ์šฐ ii. ๊ฒฝ์ œ์ ์ •์ฑ…๋ณ€์ˆ˜ x ์˜๋ณ€ํ™”๊ฐ€๊ฒฝ์ œ์ ๊ฒฐ๊ณผ y, y +, y +, y +3 ๋“ฑ์—์˜ํ–ฅ์„๋ฏธ์นจ iii. ์ด๋Š”๋‹ค์‹œ๋งํ•˜๋ฉด, y ๊ฐ€ x, x -, x

More information

ASETAOOOCRKG.hwp

ASETAOOOCRKG.hwp ์ฒญ๋…„์ธต ํฌ๋ง ์ผ์ž๋ฆฌ์™€ ์‹ค์ œ ์ทจ์—… ์ผ์ž๋ฆฌ ๊ฒฉ์ฐจ ๋ถ„์„ - ๊ณ ํ•™๋ ฅ ์ฒญ๋…„ ์‹ค์—… ์›์ธ์— ๋Œ€ํ•œ ์ผ๊ณ ์ฐฐ - ํ™ ์„ฑ ๋ฏผ * ** ๋ฐ• ์ง„ ํฌ ์„ธ๊ณ„์ ์ธ ๊ฒฝ๊ธฐ์นจ์ฒด๊ฐ€ ๋ณธ๊ฒฉํ™”๋˜๋Š” 2009๋…„์—๋Š” ์‹ค์—…๋ฌธ์ œ๊ฐ€ ๊ฐ€์žฅ ํฐ ์‚ฌํšŒ๊ฒฝ์ œ์  ์ด์Šˆ๋กœ ๋“ฑ์žฅํ•  ๊ฐ€ ๋Šฅ์„ฑ์ด ๋†’์œผ๋ฉฐ, ํŠนํžˆ ์ฒญ๋…„์ธต์˜ ๊ณ ์‹ค์—… ๋ฌธ์ œ์™€ ๋”๋ถˆ์–ด ์ผ์ž๋ฆฌ ๊ธฐํ”ผ ์ธํ•ด ๋‚˜ํƒ€๋‚  ๊ฐ€๋Šฅ์„ฑ์ด ์žˆ๋Š” NEET ํ™” ํ˜„์ƒ์— ๋Œ€ํ•œ ์šฐ๋ ค๊ฐ€ ์ปค์งˆ ๊ฒƒ์œผ๋กœ ์˜ˆ์ƒ๋œ๋‹ค.

More information

โ… . ๊ธ€๋กœ๋ฒŒ ๊ฒฝ์ œํ™˜๊ฒฝ ๋ณ€ํ™” 29๋…„์€ ์„ธ๊ณ„ ๋ฐ ๊ตญ๋‚ด๊ฒฝ์ œ์˜ ์—ญ์‚ฌ ์†์—์„œ ์˜๋ฏธ ์žˆ๋Š” ํ•œ ํ•ด๋กœ ๊ธฐ๋ก๋  ๊ฒƒ์ด๋‹ค. ๋ฆฌ๋จผ ์‡ผํฌ ์ดํ›„์˜ ๊ธˆ์œต์‹œ์žฅ ํ˜ผ๋ž€๊ณผ ๊ฒฝ์ œ์ฃผ์ฒด๋“ค์˜ ์‹ฌ๋ฆฌ ์œ„์ถ•์€ ์ƒ๋‹นํ•œ ๊ธฐ๊ฐ„์˜ ๊ฒฝ๊ธฐ๋ถˆํ™ฉ์„ ์˜ˆ๊ณ ํ•˜๋Š” ๋“ฏ ํ–ˆ์ง€๋งŒ ๊ธ€๋กœ๋ฒŒ ๊ฒฝ์ œ์œ„๊ธฐ๋ฅผ ์ˆ˜์Šตํ•˜๊ธฐ ์œ„ํ•œ ๊ฐ๊ตญ์˜ ๊ธˆ์œต์•ˆ์ •ํ™” ๋Œ€์ฑ…๊ณผ ์žฌ์ •

โ… . ๊ธ€๋กœ๋ฒŒ ๊ฒฝ์ œํ™˜๊ฒฝ ๋ณ€ํ™” 29๋…„์€ ์„ธ๊ณ„ ๋ฐ ๊ตญ๋‚ด๊ฒฝ์ œ์˜ ์—ญ์‚ฌ ์†์—์„œ ์˜๋ฏธ ์žˆ๋Š” ํ•œ ํ•ด๋กœ ๊ธฐ๋ก๋  ๊ฒƒ์ด๋‹ค. ๋ฆฌ๋จผ ์‡ผํฌ ์ดํ›„์˜ ๊ธˆ์œต์‹œ์žฅ ํ˜ผ๋ž€๊ณผ ๊ฒฝ์ œ์ฃผ์ฒด๋“ค์˜ ์‹ฌ๋ฆฌ ์œ„์ถ•์€ ์ƒ๋‹นํ•œ ๊ธฐ๊ฐ„์˜ ๊ฒฝ๊ธฐ๋ถˆํ™ฉ์„ ์˜ˆ๊ณ ํ•˜๋Š” ๋“ฏ ํ–ˆ์ง€๋งŒ ๊ธ€๋กœ๋ฒŒ ๊ฒฝ์ œ์œ„๊ธฐ๋ฅผ ์ˆ˜์Šตํ•˜๊ธฐ ์œ„ํ•œ ๊ฐ๊ตญ์˜ ๊ธˆ์œต์•ˆ์ •ํ™” ๋Œ€์ฑ…๊ณผ ์žฌ์ • 21๋…„ ๊ตญ๋‚ด๊ฒฝ์ œ ์ „๋ง ๊ฒฝ์ œ์—ฐ๊ตฌ์‹ค gtlee@lgeri.com ๊ธˆ์œต์—ฐ๊ตฌ์‹ค hybae@lgeri.com โ… . ๊ธ€๋กœ๋ฒŒ ๊ฒฝ์ œํ™˜๊ฒฝ ๋ณ€ํ™” โ…ก. ๊ตญ๋‚ด๊ฒฝ์ œ ์ „๋ง โ…ข. ๋งบ์Œ๋ง 21๋…„ ์šฐ๋ฆฌ๊ฒฝ์ œ๋Š” ์ˆ˜์š”ํšŒ๋ณต๊ณผ ๊ธฐ์ €ํšจ๊ณผ์— ํž˜์ž…์–ด ์ƒ๋ฐ˜๊ธฐ์—๋Š” 5.8% ์„ฑ์žฅ๋ฅ ์„ ๋ณด์ผ ๊ฒƒ์œผ๋กœ ์˜ˆ ์ƒ๋œ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ๊ธฐ์ €ํšจ๊ณผ๊ฐ€ ์ค„์–ด๋“œ๋Š” 21๋…„ ํ•˜๋ฐ˜๊ธฐ ์„ฑ์žฅ๋ฅ ์€ 3%๋Œ€์— ๋จธ๋ฌผ ๊ฒƒ์œผ๋กœ ๋ณด์—ฌ 21๋…„ ๊ตญ๋‚ด๊ฒฝ์ œ ์„ฑ์žฅ๋ฅ ์€

More information

<B0E6C1A6C0A7B1E25FB9DFC1A65FB0ADBCAEC8C62E687770>

<B0E6C1A6C0A7B1E25FB9DFC1A65FB0ADBCAEC8C62E687770> < ์ฃผ์ œ๋ฐœํ‘œ> ๊ณ ์กฐ๋˜๋Š” ๊ฒฝ์ œ์œ„๊ธฐ์™€ ์šฐ๋ฆฌ์˜ ๋Œ€์‘๋ฐฉ์•ˆ ๊ฐ• ์„ ํ›ˆ ( ์„ฑ์‹ ์—ฌ๋Œ€ ๊ฒฝ์ œํ•™๊ณผ ๊ต์ˆ˜) 1. ์„œ๋ก  ์„œ๋ธŒํ”„๋ผ์ž„ ์‚ฌํƒœ๋กœ ์ด‰๋ฐœ๋œ ๋ฏธ๊ตญ ๋ฐœ ์„ธ๊ณ„๊ฒฝ์ œ์˜ ์„ฑ์žฅ๋‘”ํ™”์™€ ์œ ๊ฐ€, ์›์ž์žฌ, ๊ณก๋ฌผ๊ฐ€๊ฒฉ์˜ ์ƒ์Šน์œผ๋กœ ๋‚˜ํƒ€๋‚˜๋Š” ๊ธ€๋กœ๋ฒŒ์ธํ”Œ๋ ˆ์ด์…˜ ์šฐ๋ ค์— ๋”ฐ๋ผ ์„ธ๊ณ„๊ฒฝ์ œ์˜ ์นจ์ฒด๊ฐ€๋Šฅ์„ฑ์ด ์ฆ๋Œ€๋˜๋Š” ๋™์‹œ์— ํ–ฅํ›„ ์„ธ๊ณ„ ๊ฒฝ์ œ์— ๋Œ€ํ•œ ๋ถˆํ™•์‹ค์„ฑ๋„ ์ฆํญ๋˜๊ณ  ์žˆ์Œ. ๊ตญ๋‚ด์ ์œผ๋กœ๋Š” ๊ฒฝ์ œ ์‚ด๋ฆฌ๊ธฐ๋ผ๋Š” ๊ธฐ๋Œ€๋ฅผ ๋ฐ›๊ณ 

More information

2014๋…„ ๊ฒฝ๋‚จ ๊ฒฝ์ œ์ „๋ง๊ณผ ๊ณผ์ œ 2014. 1 ์—ฐ๊ตฌ์ฑ…์ž„ ์ •๋Œ€์ฒ (๊ฒฝ๋‚จ๋ฐœ์ „์—ฐ๊ตฌ์› ์—ฐ๊ตฌ์œ„์›) ๊น€์˜์ˆœ(๊ฒฝ๋‚จ๋ฐœ์ „์—ฐ๊ตฌ์› ์—ฐ๊ตฌ์œ„์›) ์—ฐ๊ตฌ์ง€์› ๊ธธ์ˆ˜๋ฏผ(๊ฒฝ๋‚จ๋ฐœ์ „์—ฐ๊ตฌ์› ์ „๋ฌธ์—ฐ๊ตฌ์›) ๊น€๋„ํ˜•(๊ฒฝ๋‚จ๋ฐœ์ „์—ฐ๊ตฌ์› ์ „๋ฌธ์—ฐ๊ตฌ์›) ์—ฐ๊ตฌ์ž๋ฌธ ์†ํ˜•๊ทœ(ํ•œ๊ตญ์‚ฐ์—…๋‹จ์ง€๊ณต๋‹จ ๋™๋‚จ๊ถŒ๋ณธ๋ถ€ ํŒ€์žฅ) ์กฐ๋ถ€์‹(๊ฒฝ๋‚จ์ง€๋ฐฉ์ค‘์†Œ๊ธฐ์—…์ฒญ ์ฐฝ์—…์„ฑ์žฅ์ง€์›๊ณผ ํŒ€์žฅ) ์ • ํƒ€(๊ฒฝ๋‚จ์€ํ–‰ ๊ฒฝ์ œ๊ธˆ์œต์—ฐ๊ตฌ์‹ค ์ฑ…์ž„์—ฐ๊ตฌ์›) < ์š” ์•ฝ > 2014๋…„

More information

SW 2015. 02 5-1 89

SW 2015. 02 5-1 89 SW 2015. 02 88 SW 2015. 02 5-1 89 SW 2015. 02 5-2 5-3 90 SW 2015. 02 5-4 91 SW 2015. 02 5-5 5-6 92 5-7 SW 2015. 02 93 SW 2015. 02 5-8 5-1 94 SW 2015. 02 5-9 95 SW 2015. 02 5-10 5-2 96 SW 2015. 02 5-11

More information

1 SW 2015. 02 26

1 SW 2015. 02 26 02 1 SW 2015. 02 26 2-1 SW 2015. 02 27 SW 2015. 02 2-1 28 SW 2015. 02 29 2 SW 2015. 02 2-2 30 2-2 SW 2015. 02 31 SW 2015. 02 32 2-3 SW 2015. 02 33 3 SW 2015. 02 2-3 34 2-4 SW 2015. 02 35 4 SW 2015. 02

More information

201501_๊ฒฝ์ œ์ „๋ง๋ณด๊ณ ์„œ_f.hwp

201501_๊ฒฝ์ œ์ „๋ง๋ณด๊ณ ์„œ_f.hwp ISSN 2288-7083 2015. 1 2015. 1 ์š” ์•ฝ ๊ธˆ๋…„์ค‘ ๊ฒฝ์ œ์„ฑ์žฅ๋ฅ ์€ 3% ์ค‘๋ฐ˜ ์ˆ˜์ค€์„ ๋‚˜ํƒ€๋‚ผ ๊ฒƒ์œผ๋กœ ์ „๋ง ใ…‡ ๋ฏธ๊ตญ ๋“ฑ์„ ์ค‘์‹ฌ์œผ๋กœ ํ•œ ๊ธ€๋กœ๋ฒŒ ๊ฒฝ๊ธฐ ๊ฐœ์„ , ๊ตญ์ œ์œ ๊ฐ€ ํ•˜๋ฝ ๋“ฑ ์— ํž˜์ž…์–ด ์„ฑ์žฅ๋ฅ ์ด ์ ์ฐจ ๋†’์•„์งˆ ๊ฒƒ์œผ๋กœ ์˜ˆ์ƒ ์„ฑ์žฅ์— ๋Œ€ํ•œ ์ง€์ถœ๋ถ€๋ฌธ๋ณ„ ๊ธฐ์—ฌ๋„๋ฅผ ๋ณด๋ฉด ๋‚ด์ˆ˜ ๊ธฐ์—ฌ๋„๊ฐ€ ์ˆ˜์ถœ ๊ธฐ์—ฌ ๋„๋ฅผ ์ƒํšŒํ•  ์ „๋ง ์ทจ์—…์ž์ˆ˜๋Š” ๊ธˆ๋…„์ค‘ 42๋งŒ๋ช… ๋Š˜์–ด๋‚˜ 10์›” ์ „๋ง(45๋งŒ๋ช…)์—

More information

์‚ฌ์šฉ์ž ์„ค๋ช…์„œ SERVO DRIVE (FARA-CSD,CSDP-XX)

์‚ฌ์šฉ์ž ์„ค๋ช…์„œ SERVO DRIVE (FARA-CSD,CSDP-XX) A3 : 30W 10 : 1kW A5 : 50W 15 : 1.5kW 01 : 100W 20 : 2kW.... 08 : 800W 50 : 5kW CSD series CSDD series CSDF series CSDS series CSDH series CSDN series CSDX series A : AC110V B : AC220V S : / P : 20?C

More information

์š” ์•ฝ ๋Œ€ํ•œ๋ฏผ๊ตญ์€ 1948๋…„ ์ •๋ถ€์ˆ˜๋ฆฝ ์ดํ›„ 60๋…„ ๋งŒ์— ์„ธ๊ณ„ 13์œ„์˜ ๊ฒฝ์ œ๋Œ€๊ตญ์œผ๋กœ ๋น„์•ฝ์ ์ธ ์„ฑ ์žฅ์„ ์ด๋ฃจ์—ˆ๋‹ค. ์ •๋ถ€์ˆ˜๋ฆฝ 2๋…„ ๋งŒ์— ๋ถํ•œ์˜ ์ „๊ฒฉ ๋‚จ์นจ์œผ๋กœ ์ „ ๊ตญํ† ๊ฐ€ ์ดˆํ† ํ™”๋˜์—ˆ๊ณ , ํœด ์ „ ์ดํ›„์—๋„ ์•ˆ๋ณด์— ๋Œ€ํ•œ ์œ„ํ˜‘์€ ๊ณ„์†๋˜์—ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ๋Œ€ํ•œ๋ฏผ๊ตญ ๊ตญ๋ฏผ์€ ๋ถˆ๋ฆฌํ•œ ์—ฌ๊ฑด์— ์ขŒ์ ˆํ•˜

์š” ์•ฝ ๋Œ€ํ•œ๋ฏผ๊ตญ์€ 1948๋…„ ์ •๋ถ€์ˆ˜๋ฆฝ ์ดํ›„ 60๋…„ ๋งŒ์— ์„ธ๊ณ„ 13์œ„์˜ ๊ฒฝ์ œ๋Œ€๊ตญ์œผ๋กœ ๋น„์•ฝ์ ์ธ ์„ฑ ์žฅ์„ ์ด๋ฃจ์—ˆ๋‹ค. ์ •๋ถ€์ˆ˜๋ฆฝ 2๋…„ ๋งŒ์— ๋ถํ•œ์˜ ์ „๊ฒฉ ๋‚จ์นจ์œผ๋กœ ์ „ ๊ตญํ† ๊ฐ€ ์ดˆํ† ํ™”๋˜์—ˆ๊ณ , ํœด ์ „ ์ดํ›„์—๋„ ์•ˆ๋ณด์— ๋Œ€ํ•œ ์œ„ํ˜‘์€ ๊ณ„์†๋˜์—ˆ๋‹ค. ๊ทธ๋Ÿฌ๋‚˜ ๋Œ€ํ•œ๋ฏผ๊ตญ ๊ตญ๋ฏผ์€ ๋ถˆ๋ฆฌํ•œ ์—ฌ๊ฑด์— ์ขŒ์ ˆํ•˜ 2008.8.13. (์ œ667ํ˜ธ) ๋Œ€ํ•œ๋ฏผ๊ตญ๊ฒฝ์ œ 60๋…„์˜ ๅคง ้•ท ็จ‹ ๋ชฉ์ฐจ โ… . ๋Œ€ํ•œ๋ฏผ๊ตญ๊ฒฝ์ œ 60๋…„์˜ ์„ฑ๊ณผ โ…ก. ๋Œ€ํ•œ๋ฏผ๊ตญ๊ฒฝ์ œ์˜ ์—ฌ๊ฑด๊ณผ ์ œ๋„ โ…ข. ๋Œ€ํ•œ๋ฏผ๊ตญ๊ฒฝ์ œ์˜ ์ •์ฑ… ์„ ํƒ โ…ฃ. ํ•œ๊ตญ๊ฒฝ์ œ์˜ ๋‹น๋ฉด๊ณผ์ œ ์ž‘์„ฑ : ์ „์˜์žฌ ์ˆ˜์„์—ฐ๊ตฌ์›(3780-8357) serijyj@seri.org ํ™ฉ์ธ์„ฑ, ๊น€๋“๊ฐ‘, ์žฅ์„ฑ์›, ํ•œ์ฐฝ์ˆ˜, ์ „ํšจ์ฐฌ, ๊ฐ•์„ฑ์›, ์ž„์ˆ˜ํ˜ธ, ์†๋ฏผ์ค‘ ๊ฐ์ˆ˜ : ๊ถŒ์ˆœ์šฐ ์ˆ˜์„์—ฐ๊ตฌ์›(3780-8071)

More information

, ( ) * 1) *** *** (KCGS) 2003, 2004 (CGI),. (+),.,,,.,. (endogeneity) (reverse causality),.,,,. I ( ) *. ** ***

, ( ) * 1) *** *** (KCGS) 2003, 2004 (CGI),. (+),.,,,.,. (endogeneity) (reverse causality),.,,,. I ( ) *. ** *** , 40 3 4 (2006 12 ) * 1) *** *** (KCGS) 2003, 2004 (CGI),. (+),.,,,.,. (endogeneity) (reverse causality),.,,,. I. 1998. 2005 12 ( ) *. ** *** 2, 40 3 4 37.2%, 20 60%. 80%..,..,.,,, (SCB),,,.,..,, /,..

More information

็ฌฌ 1 ็ฏ€ ็ต„ ็น” 11 ็ฌฌ 1 ็ซ  ๆชข ๅฏŸ ์˜ ็ต„ ็น” ไบบ ไบ‹ ๅˆถ ๅบฆ ๋“ฑ ็ฌฌ 1 ้ … ๅคง ๆชข ๅฏŸ ๅปณ ็ฌฌ 1 ็ฏ€ ็ต„ ๋Œ€๊ฒ€์ฐฐ์ฒญ์€ ๋Œ€๋ฒ•์›์— ๋Œ€์‘ํ•˜์—ฌ ์ˆ˜๋„์ธ ์„œ์šธ์— ์œ„์น˜ ํ•œ๋‹ค(๊ฒ€์ฐฐ์ฒญ๋ฒ• ์ œ2์กฐ,์ œ3์กฐ,๋Œ€๊ฒ€์ฐฐ์ฒญ์˜ ์œ„์น˜์™€ ๊ฐ๊ธ‰ ๊ฒ€์ฐฐ์ฒญ์˜๋ช…์นญ๋ฐ์œ„์น˜์—๊ด€ํ•œ๊ทœ์ • ์ œ2์กฐ). ๋Œ€๊ฒ€์ฐฐ์ฒญ์— ๊ฒ€์ฐฐ์ด์žฅ,๋Œ€

็ฌฌ 1 ็ฏ€ ็ต„ ็น” 11 ็ฌฌ 1 ็ซ  ๆชข ๅฏŸ ์˜ ็ต„ ็น” ไบบ ไบ‹ ๅˆถ ๅบฆ ๋“ฑ ็ฌฌ 1 ้ … ๅคง ๆชข ๅฏŸ ๅปณ ็ฌฌ 1 ็ฏ€ ็ต„ ๋Œ€๊ฒ€์ฐฐ์ฒญ์€ ๋Œ€๋ฒ•์›์— ๋Œ€์‘ํ•˜์—ฌ ์ˆ˜๋„์ธ ์„œ์šธ์— ์œ„์น˜ ํ•œ๋‹ค(๊ฒ€์ฐฐ์ฒญ๋ฒ• ์ œ2์กฐ,์ œ3์กฐ,๋Œ€๊ฒ€์ฐฐ์ฒญ์˜ ์œ„์น˜์™€ ๊ฐ๊ธ‰ ๊ฒ€์ฐฐ์ฒญ์˜๋ช…์นญ๋ฐ์œ„์น˜์—๊ด€ํ•œ๊ทœ์ • ์ œ2์กฐ). ๋Œ€๊ฒ€์ฐฐ์ฒญ์— ๊ฒ€์ฐฐ์ด์žฅ,๋Œ€ ็ฌฌ 1 ็ซ  ๆชข ๅฏŸ ์˜ ็ต„ ็น” ไบบ ไบ‹ ๅˆถ ๅบฆ ๋“ฑ ็ฌฌ 1 ็ฏ€ ็ต„ ็น” 11 ็ฌฌ 1 ็ซ  ๆชข ๅฏŸ ์˜ ็ต„ ็น” ไบบ ไบ‹ ๅˆถ ๅบฆ ๋“ฑ ็ฌฌ 1 ้ … ๅคง ๆชข ๅฏŸ ๅปณ ็ฌฌ 1 ็ฏ€ ็ต„ ๋Œ€๊ฒ€์ฐฐ์ฒญ์€ ๋Œ€๋ฒ•์›์— ๋Œ€์‘ํ•˜์—ฌ ์ˆ˜๋„์ธ ์„œ์šธ์— ์œ„์น˜ ํ•œ๋‹ค(๊ฒ€์ฐฐ์ฒญ๋ฒ• ์ œ2์กฐ,์ œ3์กฐ,๋Œ€๊ฒ€์ฐฐ์ฒญ์˜ ์œ„์น˜์™€ ๊ฐ๊ธ‰ ๊ฒ€์ฐฐ์ฒญ์˜๋ช…์นญ๋ฐ์œ„์น˜์—๊ด€ํ•œ๊ทœ์ • ์ œ2์กฐ). ๋Œ€๊ฒ€์ฐฐ์ฒญ์— ๊ฒ€์ฐฐ์ด์žฅ,๋Œ€๊ฒ€์ฐฐ์ฒญ ์ฐจ์žฅ๊ฒ€์‚ฌ,๋Œ€๊ฒ€์ฐฐ์ฒญ ๊ฒ€์‚ฌ,๊ฒ€์ฐฐ์—ฐ๊ตฌ๊ด€,๋ถ€

More information

13๋ฌผ๊ฐ€๋ณด๊ณ ์„œ01-๋„๋น„๋ผ๋ฐ๋ชฉ์ฐจ1~7

13๋ฌผ๊ฐ€๋ณด๊ณ ์„œ01-๋„๋น„๋ผ๋ฐ๋ชฉ์ฐจ1~7 ISSN 87-81 1. 1 1. 1 11. -...7 1....7. / 1.7 -. 1.9. 1. -. 1.7. 11 1.8 -... 1 1-1 -1 1 1 i 1 1 1 1 ii 1. 1.. 1. 1.... -. -. -1. -1. -1. -1. 1 1 1 1 1 1 1 1 11 11 1 1 9 9 8 8 7 7 iii 9 9 7 7 8 8

More information

<28C1A6313236C8A32920B9CCB1B920BEE7C0FBBFCFC8ADC1A4C3A5C0CC20B1B9B3BBB0E6C1A6BFA120B9CCC4A1B4C220BFB5C7E22E687770>

<28C1A6313236C8A32920B9CCB1B920BEE7C0FBBFCFC8ADC1A4C3A5C0CC20B1B9B3BBB0E6C1A6BFA120B9CCC4A1B4C220BFB5C7E22E687770> 2010 NHERI ๋ฆฌํฌํŠธ ์ œ126ํ˜ธ (2010. 11. 18) ์ž‘ ์„ฑ : ์œค๊ฑด์šฉ ์ˆ˜์„์—ฐ๊ตฌ์›(6399-5908) kyyun0129@daum.net ๋ฐ•ํƒœ์˜ ์ฑ…์ž„์—ฐ๊ตฌ์›(6399-5935) iusti@hotmail.com ๊ฐ ์ˆ˜ : ์ž„์ค€ํ™˜ ๊ฑฐ์‹œ๊ฒฝ์ œ์—ฐ๊ตฌ์‹ค์žฅ(6399-5981) pricing@paran.com ๋ฏธ๊ตญ ์–‘์ ์™„ํ™”์ •์ฑ…์ด ๊ตญ๋‚ด๊ฒฝ์ œ์— ๋ฏธ์น˜๋Š” ์˜ํ–ฅ ๋ชฉ ์ฐจ

More information

ๆญฏํ‘œ์ง€_์ตœ์ข…H_.PDF

ๆญฏํ‘œ์ง€_์ตœ์ข…H_.PDF 21 5 LG < > / 5 5 Chen, Roll, Ross,, (-) (+) (-), (-) (+) / (-), (+) J, /, (-) IMF, /,,, 5bp 5bp 1 1 3 ( )/ ( ) ( :p) 1 3 6 / 1-23 45 55 1% 2 33 42 1p 296 234 175 1%p -13 147 171 1%p 1 81 7 1% 48 57 54

More information

Precipitation prediction of numerical analysis for Mg-Al alloys

Precipitation prediction of numerical analysis for Mg-Al alloys ์ €์ž‘์žํ‘œ์‹œ - ๋น„์˜๋ฆฌ - ๋ณ€๊ฒฝ๊ธˆ์ง€ 2.0 ๋Œ€ํ•œ๋ฏผ๊ตญ ์ด์šฉ์ž๋Š”์•„๋ž˜์˜์กฐ๊ฑด์„๋”ฐ๋ฅด๋Š”๊ฒฝ์šฐ์—ํ•œํ•˜์—ฌ์ž์œ ๋กญ๊ฒŒ ์ด์ €์ž‘๋ฌผ์„๋ณต์ œ, ๋ฐฐํฌ, ์ „์†ก, ์ „์‹œ, ๊ณต์—ฐ๋ฐ๋ฐฉ์†กํ• ์ˆ˜์žˆ์Šต๋‹ˆ๋‹ค. ๋‹ค์Œ๊ณผ๊ฐ™์€์กฐ๊ฑด์„๋”ฐ๋ผ์•ผํ•ฉ๋‹ˆ๋‹ค : ์ €์ž‘์žํ‘œ์‹œ. ๊ท€ํ•˜๋Š”์›์ €์ž‘์ž๋ฅผํ‘œ์‹œํ•˜์—ฌ์•ผํ•ฉ๋‹ˆ๋‹ค. ๋น„์˜๋ฆฌ. ๊ท€ํ•˜๋Š”์ด์ €์ž‘๋ฌผ์„์˜๋ฆฌ๋ชฉ์ ์œผ๋กœ์ด์šฉํ• ์ˆ˜์—†์Šต๋‹ˆ๋‹ค. ๋ณ€๊ฒฝ๊ธˆ์ง€. ๊ท€ํ•˜๋Š”์ด์ €์ž‘๋ฌผ์„๊ฐœ์ž‘, ๋ณ€ํ˜•๋˜๋Š”๊ฐ€๊ณตํ• ์ˆ˜์—†์Šต๋‹ˆ๋‹ค. ๊ท€ํ•˜๋Š”, ์ด์ €์ž‘๋ฌผ์˜์žฌ์ด์šฉ์ด๋‚˜๋ฐฐํฌ์˜๊ฒฝ์šฐ,

More information

ๆญฏํ‘œ์ง€.PDF

ๆญฏํ‘œ์ง€.PDF 2000 11 ,, LG i < > 1023 44 40 1999 2,497 ( 479 ) 600900, 500600, 50% (4 ) 146% 4 10%, 25% 4 05% ( 10 ) 86,646, 122,489, 50% ILO 25%, 1 : 2 : 2001 3 : 2001 40 2010 (GDP )34% 320012003 ii GDP,,, ( 98 4/4

More information

ํŠน์ง‘....,.,., (good jobs) (rent-sharing) (fairness)..... โ…ก. ์ž„๊ธˆ๊ณผ์ƒ์‚ฐ์„ฑ๊ตฌ์กฐ์˜๋ถ„์„๋ชจํ˜• ) 1),,,, 2_ ๋…ธ๋™๋ฆฌ๋ทฐ

ํŠน์ง‘....,.,., (good jobs) (rent-sharing) (fairness)..... โ…ก. ์ž„๊ธˆ๊ณผ์ƒ์‚ฐ์„ฑ๊ตฌ์กฐ์˜๋ถ„์„๋ชจํ˜• ) 1),,,, 2_ ๋…ธ๋™๋ฆฌ๋ทฐ ์—ฐ๊ณต์ž„๊ธˆ์„๋‹ค์‹œ์ƒ๊ฐํ•œ๋‹ค ์—ฐ๊ณต์ž„๊ธˆ์„๋‹ค์‹œ์ƒ๊ฐํ•œ๋‹ค 1) ํ™ฉ์ˆ˜๊ฒฝ * โ… . ๋จธ๋ฆฌ๋ง ().......... * (skhwang@kli.re.kr). ํŠน์ง‘ _1 ํŠน์ง‘....,.,., (good jobs) (rent-sharing) (fairness)..... โ…ก. ์ž„๊ธˆ๊ณผ์ƒ์‚ฐ์„ฑ๊ตฌ์กฐ์˜๋ถ„์„๋ชจํ˜•. 1980 1) 1),,,, 2_ ๋…ธ๋™๋ฆฌ๋ทฐ ์—ฐ๊ณต์ž„๊ธˆ์„๋‹ค์‹œ์ƒ๊ฐํ•œ๋‹ค 1990 -. Hellerstein

More information

Microsoft Word - MacroVision_121105_.doc

Microsoft Word - MacroVision_121105_.doc Weekly Focus ์›/๋‹ฌ๋Ÿฌ ํ™˜์œจ ๋ณ€๋™์˜ ๊ฑฐ์‹œ๊ฒฝ์ œ ํŒŒ๊ธ‰ํšจ๊ณผ์™€ ์ „๋ง Economist ๊น€์œค๊ธฐ (769-3063) mackyg@deri.co.kr Summary ์ตœ๊ทผ ์›/๋‹ฌ๋Ÿฌ ํ™˜์œจ ํ•˜๋ฝ์ด ๋น ๋ฅด๊ฒŒ ์ง„ํ–‰๋˜๋ฉด์„œ ์šฐ๋ฆฌ ๊ฒฝ์ œ์— ๋ฏธ์น  ๋ถ€์ •์  ์˜ํ–ฅ์— ๋Œ€ํ•œ ์šฐ๋ ค๊ฐ€ ์ฆ๋Œ€ํ•˜๊ณ  ์žˆ์Œ. ์›/๋‹ฌ๋Ÿฌ ํ™˜์œจ์€ 10/29์ผ ์‹ฌ๋ฆฌ์  ๋งˆ์ง€๋…ธ์„ ์œผ๋กœ ์—ฌ๊ฒจ์ง€๋˜ 1,100์›์„ ์ด ๋ถ•๊ดด๋˜ ์—ˆ์œผ๋ฉฐ 11/2์ผ

More information

- 2 -

- 2 - - 1 - - 2 - - 3 - - 4 - - 1 - - 2 - ๊ตฌ๋ถ„์ฒญ๊ตฌ ์‹ฌ๊ฒฐ (B) ์ทจํ•˜์ ˆ์ฐจ๋ฌดํšจ๋ฏธ์ฒ˜๋ฆฌ (A) ์ธ์šฉ๊ธฐ๊ฐ๊ฐํ•˜์†Œ๊ณ„ (C) (D) (E=A-(B+C+D) 2015 505 0 165 0 165 176 116 48 2016 3 0 2 0 2 0 0 1 ํ•ฉ๊ณ„ 508 0 167 0 167 176 116 49 ๊ตฌ๋ถ„ ์‹ฌ๊ฒฐ๋…„ 2013 2014 2015 2016

More information

CPI CORE CPI 125 115 105 95 (04.01.02 = 100) 85 75 04.01 04.07 05.01 05.07 06.01 15% 10% 5% 0% -5% 96.01 98.01 00.01 02.01 04.01 06.01

CPI CORE CPI 125 115 105 95 (04.01.02 = 100) 85 75 04.01 04.07 05.01 05.07 06.01 15% 10% 5% 0% -5% 96.01 98.01 00.01 02.01 04.01 06.01 Best Investment Partner CPI CORE CPI 125 115 105 95 (04.01.02 = 100) 85 75 04.01 04.07 05.01 05.07 06.01 15% 10% 5% 0% -5% 96.01 98.01 00.01 02.01 04.01 06.01 LG LCD P7., 87.5% 88.6%. 1/4 740 17.4%. T/K

More information

IDP www idp or kr IDP ์ •์ฑ…์—ฐ๊ตฌ ํ•œ๊ตญ๊ฒฝ์ œ์˜๊ตฌ์กฐ์ ๋ฌธ์ œ์™€๊ฐœํ˜๋ฐฉํ–ฅ ๋ฏผ์ฃผ์ •์ฑ…์—ฐ๊ตฌ์› The Institute for Democracy and Policies

IDP www idp or kr IDP ์ •์ฑ…์—ฐ๊ตฌ ํ•œ๊ตญ๊ฒฝ์ œ์˜๊ตฌ์กฐ์ ๋ฌธ์ œ์™€๊ฐœํ˜๋ฐฉํ–ฅ ๋ฏผ์ฃผ์ •์ฑ…์—ฐ๊ตฌ์› The Institute for Democracy and Policies IDP ์ •์ฑ…์—ฐ๊ตฌ 2011-13 www.idp.or.kr IDP ์ •์ฑ…์—ฐ๊ตฌ 2011-13 ํ•œ๊ตญ๊ฒฝ์ œ์˜ ๊ตฌ์กฐ์  ๋ฌธ์ œ์™€ ๊ฐœํ˜๋ฐฉํ–ฅ The Institute for Democracy and Policies ์„œ์šธ์‹œ ์˜๋“ฑํฌ๊ตฌ ์˜๋“ฑํฌ๋™ 6๊ฐ€ 133๋ฒˆ์ง€ ๋ฏผ์ฃผ์ •์ฑ…์—ฐ๊ตฌ์› Tel : 02.2630.0114 Fax : 02.2630.0194 ๋ฏผ์ฃผ์ •์ฑ…์—ฐ๊ตฌ์› ๋ฏผ์ฃผ์ •์ฑ…์—ฐ๊ตฌ์› IDP ์ •์ฑ…์—ฐ๊ตฌ

More information

LG Business Insight 1377

LG Business Insight 1377 ์ค‘๊ตญ ๊ฒฝ์ œ์˜ ๊ฒฝ์ฐฉ๋ฅ™ ๊ฐ€๋Šฅ์„ฑ๊ณผ ์˜ํ–ฅ ์ด์ฒ ์šฉ ์—ฐ๊ตฌ์œ„์› lcy@lgeri.com 1. ์ค‘๊ตญ ๊ฒฝ์ œ ๊ฒฝ์ฐฉ๋ฅ™์— ๋Œ€ํ•œ ์šฐ๋ ค ๊ณ ์กฐ 2. ๊ฒฝ์ฐฉ๋ฅ™์„ ์œ ๋ฐœํ•  ์ˆ˜ ์žˆ๋Š” ์š”์ธ๋“ค 3. ํ–ฅํ›„ ์ค‘๊ตญ ๊ฒฝ์ œ์˜ ์„ธ ๊ฐ€์ง€ ์ง„๋กœ 4. ์ค‘๊ตญ ๊ฒฝ์ œ๊ฐ€ ๊ฒฝ์ฐฉ๋ฅ™ ํ•œ๋‹ค๋ฉด... ์˜ฌํ•ด ๋“ค์–ด ์„ฑ์žฅ ๋‘”ํ™” ์ง€์†, ๊ธ‰๊ฒฉํ•œ ์ฃผ๊ฐ€ ์กฐ์ •, ๊ธฐ์Šต์ ์ธ ์œ„์•ˆํ™” ํ‰๊ฐ€์ ˆํ•˜, ์ž‡๋”ฐ๋ฅธ ๊ธฐ์—… ๋ถ€๋„ ๋“ฑ ๋ถ€์ •์ ์ธ ๋‰ด์Šค๊ฐ€ ์Ÿ์•„์ง€๋ฉด์„œ ์ค‘๊ตญ ๊ฒฝ์ œ

More information

ๅผบ ๋‹ฌ๋Ÿฌ ํ˜„์ƒ์ด ์‹ ํฅ๊ตญ ๊ฒฝ์ œ์— ๋ฏธ์น˜๋Š” ์˜ํ–ฅ ๋ฐ ๋Œ€์‘๋ฐฉ์•ˆ 1. ์ตœ๊ทผ ๅผบ ๋‹ฌ๋Ÿฌ ํ˜„์ƒ ๋ฐฐ๊ฒฝ ๋ฐ ๊ฒฝ๊ณผ ๋ฐœ์ƒ ๋ฐฐ๊ฒฝ ์–‘์  ์™„ํ™” ๊ธฐ๊ฐ„์ค‘ ํ•ด์™ธ๋กœ ํˆฌ์ž๋œ ๋‹ฌ๋Ÿฌํ™”๊ฐ€ ๋ฏธ๊ตญ์˜ ํ…Œ์ดํผ๋ง ์ •์ฑ…๊ณผ ๊ธˆ๋ฆฌ ์ธ ์ƒ ์ „๋ง์— ๋”ฐ๋ผ ๋‹ค์‹œ ๋ฏธ๊ตญ์œผ๋กœ ์œ ์ž…๋˜๋ฉด์„œ ํ•ด์™ธ ๊ฐ ๊ตญ์˜ ๋‹ฌ๋Ÿฌํ™” ์ˆ˜์š”์— ๋น„ํ•ด ๊ณต๊ธ‰ ๋ถ€์กฑ์œผ๋กœ

ๅผบ ๋‹ฌ๋Ÿฌ ํ˜„์ƒ์ด ์‹ ํฅ๊ตญ ๊ฒฝ์ œ์— ๋ฏธ์น˜๋Š” ์˜ํ–ฅ ๋ฐ ๋Œ€์‘๋ฐฉ์•ˆ 1. ์ตœ๊ทผ ๅผบ ๋‹ฌ๋Ÿฌ ํ˜„์ƒ ๋ฐฐ๊ฒฝ ๋ฐ ๊ฒฝ๊ณผ ๋ฐœ์ƒ ๋ฐฐ๊ฒฝ ์–‘์  ์™„ํ™” ๊ธฐ๊ฐ„์ค‘ ํ•ด์™ธ๋กœ ํˆฌ์ž๋œ ๋‹ฌ๋Ÿฌํ™”๊ฐ€ ๋ฏธ๊ตญ์˜ ํ…Œ์ดํผ๋ง ์ •์ฑ…๊ณผ ๊ธˆ๋ฆฌ ์ธ ์ƒ ์ „๋ง์— ๋”ฐ๋ผ ๋‹ค์‹œ ๋ฏธ๊ตญ์œผ๋กœ ์œ ์ž…๋˜๋ฉด์„œ ํ•ด์™ธ ๊ฐ ๊ตญ์˜ ๋‹ฌ๋Ÿฌํ™” ์ˆ˜์š”์— ๋น„ํ•ด ๊ณต๊ธ‰ ๋ถ€์กฑ์œผ๋กœ Global Issue Paper 2014. 10.28 ์„œ์šธ์‹œ ์„œ์ดˆ๊ตฌ ํ—Œ๋ฆ‰๋กœ 13 TEL 02-3460-7314 FAX 02 3460 7777 ๊ธ€๋กœ๋ฒŒ ์ด์Šˆ ํŽ˜์ดํผ ์ •๋ณด๋Š” ์ „ ์„ธ๊ณ„ 124๊ฐœ ๋ฌด์—ญ๊ด€์—์„œ ์˜ค๋Š” ์‹ค๋ฌผ ๊ฒฝ์ œ์ •๋ณด๋ฅผ ์ทจํ•ฉ ๋ถ„์„ํ•˜๊ณ  ํ˜„์•ˆ ์ด์Šˆ๊ฐ€ ๋ฐœ์ƒํ•œ ํ•ด์™ธ ํ˜„์žฅ์˜ ์†Œ๋ฆฌ๋ฅผ ๋ฐ˜์˜ํ•œ ๊ฒƒ์œผ๋กœ, ์šฐ๋ฆฌ๋‚˜๋ผ ๋ฌด์—ญ ํˆฌ์ž ํ™œ๋™ ์ง€์›์„ ์œ„ํ•ด ์ž‘์„ฑ๋œ ๊ฒƒ์ž…๋‹ˆ๋‹ค. ์ตœ๊ทผ ๅผบ ๋‹ฌ๋Ÿฌ

More information

. 2011 4.0% 2010 5%. 6 4.3% 0.3% 2010 6.4% 10 3.1% 1.6%. 6 2.2% 0.6%. 9.5%, 7.8% 2012 4.0%. 6 4.5% 0.5% 6.1%. 6 6.4%. 1.9%. 6 2.6% 0.7%. 2.7% 1.8%, 1.

. 2011 4.0% 2010 5%. 6 4.3% 0.3% 2010 6.4% 10 3.1% 1.6%. 6 2.2% 0.6%. 9.5%, 7.8% 2012 4.0%. 6 4.5% 0.5% 6.1%. 6 6.4%. 1.9%. 6 2.6% 0.7%. 2.7% 1.8%, 1. . 2011 4.0% 2010 5%. 6 4.3% 0.3% 2010 6.4% 10 3.1% 1.6%. 6 2.2% 0.6%. 9.5%, 7.8% 2012 4.0%. 6 4.5% 0.5% 6.1%. 6 6.4%. 1.9%. 6 2.6% 0.7%. 2.7% 1.8%, 1.7% 1.1%, 2.9% 2.3%. IMF IMF 2 3 3 4 8 5 S&P : GDP :

More information

2002.9์›”์ž‘์—….doc

2002.9์›”์ž‘์—….doc 2002 [ ] 1...... 1 2.. 2 3.. 3 4. /.. 4 5...... 5 6.. 6 7.. 7 8.. 8 9.... 9 10..... 10 11. 10 12. 11 .,, CI,. Moody s, S&P, Fitch, The Banker 3., 247,620 2138,405, 179,653 1684,656, 197,006 1267,306. 22,657,

More information

G20 ์ œ1์ฐจ ์ •์ƒํšŒ์˜ : ๋ฏธ๊ตญ ์›Œ์‹ฑํ„ด D.C.(2008. 11. 15) ์ฐธ๊ฐ€๊ตญ(๊ธฐ๊ตฌ) G20 ํšŒ์›๊ตญ : G7 ํ•œ๊ตญ ์ค‘๊ตญ ์ธ๋„ ํ˜ธ์ฃผ ๋ธŒ๋ผ์งˆ ๋ฉ•์‹œ์ฝ” ์ธ๋„๋„ค์‹œ์•„ ์•„๋ฅดํ—จํ‹ฐ๋‚˜ ๋Ÿฌ์‹œ์•„ ํ„ฐํ‚ค ์‚ฌ์šฐ๋””์•„๋ผ๋น„์•„ ๋‚จ์•„ํ”„๋ฆฌ์นด๊ณตํ™”๊ตญ ์˜์žฅ๊ตญ(ํ”„๋ž‘์Šค) ์ŠคํŽ˜์ธ ๋„ค๋œ๋ž€๋“œ ๊ตญ ์ œ ๊ธฐ ๊ตฌ : UN I

G20 ์ œ1์ฐจ ์ •์ƒํšŒ์˜ : ๋ฏธ๊ตญ ์›Œ์‹ฑํ„ด D.C.(2008. 11. 15) ์ฐธ๊ฐ€๊ตญ(๊ธฐ๊ตฌ) G20 ํšŒ์›๊ตญ : G7 ํ•œ๊ตญ ์ค‘๊ตญ ์ธ๋„ ํ˜ธ์ฃผ ๋ธŒ๋ผ์งˆ ๋ฉ•์‹œ์ฝ” ์ธ๋„๋„ค์‹œ์•„ ์•„๋ฅดํ—จํ‹ฐ๋‚˜ ๋Ÿฌ์‹œ์•„ ํ„ฐํ‚ค ์‚ฌ์šฐ๋””์•„๋ผ๋น„์•„ ๋‚จ์•„ํ”„๋ฆฌ์นด๊ณตํ™”๊ตญ ์˜์žฅ๊ตญ(ํ”„๋ž‘์Šค) ์ŠคํŽ˜์ธ ๋„ค๋œ๋ž€๋“œ ๊ตญ ์ œ ๊ธฐ ๊ตฌ : UN I G20 ์ฝ”๋ฆฌ์•„ ๋” ํฐ ์„ธ๊ณ„๋กœ G20 ์ •์ƒํšŒ์˜ ์˜๋ฏธ์™€ ๊ณผ์ œ ๊ทธ๋ฆฌ๊ณ  ๋Œ€ํ•œ๋ฏผ๊ตญ์˜ ์ƒˆ๋กœ์šด ๋„์ „ ๋ฌธํ™”์ฒด์œก๊ด€๊ด‘๋ถ€ G20 ์ •์ƒํšŒ์˜ ์ค€๋น„์œ„์›ํšŒ G20 ์ œ1์ฐจ ์ •์ƒํšŒ์˜ : ๋ฏธ๊ตญ ์›Œ์‹ฑํ„ด D.C.(2008. 11. 15) ์ฐธ๊ฐ€๊ตญ(๊ธฐ๊ตฌ) G20 ํšŒ์›๊ตญ : G7 ํ•œ๊ตญ ์ค‘๊ตญ ์ธ๋„ ํ˜ธ์ฃผ ๋ธŒ๋ผ์งˆ ๋ฉ•์‹œ์ฝ” ์ธ๋„๋„ค์‹œ์•„ ์•„๋ฅดํ—จํ‹ฐ๋‚˜ ๋Ÿฌ์‹œ์•„ ํ„ฐํ‚ค ์‚ฌ์šฐ๋””์•„๋ผ๋น„์•„ ๋‚จ์•„ํ”„๋ฆฌ์นด๊ณตํ™”๊ตญ ์˜์žฅ๊ตญ(ํ”„๋ž‘์Šค) ์ŠคํŽ˜์ธ

More information

๊ตญ์ œ๋ฌด์—ญ๋ก -02์žฅ

๊ตญ์ œ๋ฌด์—ญ๋ก -02์žฅ 1 1 ( 27 ). 2, 3, 4. - 5. 6,. 7. 2... 17 20... 32 1.? ( ).?? 2.?,? 17 18, (Adam Smith). 40 (David Ricardo).,.,,. 20 (Gottfried Haberler). 2 2. 2 2.,. 6., 7. 1776. 1776,,,., 17 18 (,, ).,. 33 (1571 1641).

More information

Microsoft Word - ํ•ซ์ด์Šˆ 0201 ์ตœ์ข…_๊ฒฐํ•ฉ_.doc

Microsoft Word - ํ•ซ์ด์Šˆ 0201 ์ตœ์ข…_๊ฒฐํ•ฉ_.doc ๊น€์žฌํ™ Hot Issue ๋”ฐ๋ผ์žก๊ธฐ Tel. -6 Tel. -6 kim.jae-hong@shinyoung.com chun.won-chang@shinyoung.com ๊ฒฝ๊ธฐ, ์œ ๋™์„ฑ์ด ๋…„์„ ๊ฐ€๋ฆฌํ‚ค๊ณ  ์žˆ๋‚˜? ์›” ์ค‘ ์œ„ํ—˜์ž์‚ฐ ์„ ํ˜ธํ˜„์ƒ์ด ์ง„ํ–‰๋˜์—ˆ๋‹ค. ์œ ๋™์„ฑ๊ณผ ๊ฒฝ๊ธฐ๊ฐœ์„ ์— ๋Œ€ํ•œ ๊ธฐ๋Œ€๋Š” ๋…„์˜ ๊ฐ•์„ธ์žฅ์„ ์—ฐ์ƒํ•˜๊ฒŒ ๋งŒ๋“ ๋‹ค. ์ด์— ๋Œ€ํ•œ ์šฐ๋ฆฌ์˜ ์‹œ๊ฐ์€ ๋‹ค์Œ๊ณผ ๊ฐ™๋‹ค. ) 8~๋…„๊ณผ ๋น„์Šทํ•œ

More information

Microsoft Word - 2016052014532042K_01_01.docx

Microsoft Word - 2016052014532042K_01_01.docx 216. 5. 2 Week Ahead: 5์›” ๋„ท์งธ์ฃผ 4์›” FOMC ํšŒ์˜๋ก: 6์›”์ด๋ƒ 7์›”์ด๋ƒ Global Macro Team Analysts ํ—ˆ์ง„์šฑ (Global/US) james_huh@samsung.com 2 22 7796 ์ด์Šนํ›ˆ (China/Korea) sh213.lee@samsung.com 2 22 7836 ๊น€์ง€์€ (Europe) jieun26.kim@samsung.com

More information

์Šฌ๋ผ์ด๋“œ 1

์Šฌ๋ผ์ด๋“œ 1 Principles of Economerics (3e) Ch. 9 ๋™ํƒœ๋ชจํ˜•, ์ž๊ธฐ์ƒ๊ด€, ์˜ˆ์ธก 2013 ๋…„ 1 ํ•™๊ธฐ ์œค์„ฑ๋ฏผ 9.1 ์„œ๋ก  ๊ฒฝ์ œ์ ์ธ๊ณผ๊ด€๊ณ„์˜์‹œ์ฐจํšจ๊ณผ x ๋ณ€ํ™”๋Š”์—ฌ๋Ÿฌ๊ธฐ๊ฐ„์—๊ฑธ์ณ์ข…์†๋ณ€์ˆ˜์—์˜ํ–ฅ์„๋ฏธ์น ์ˆ˜์žˆ์Œ x y, y+1, y+ 2, 2 9.1 ์„œ๋ก  ๊ฒฝ์ œ๋ณ€์ˆ˜์˜๋™ํƒœ์ ๊ด€๊ณ„๋ฅผ๋ชจํ˜•ํ™”ํ• ์ˆ˜์žˆ๋Š”์„ธ๊ฐ€์ง€๋ฐฉ๋ฒ• ์‹œ์ฐจ์„ค๋ช…๋ณ€์ˆ˜์˜๋„์ž… x y, y+1, y+ 2, ์ด๊ด€๊ณ„๋ฅผ๋‹ฌ๋ฆฌํ‘œํ˜„ํ•˜๋ฉด,

More information

BIS Solvency (RBC) Solvency. Solvency,. Solvency.

BIS Solvency (RBC) Solvency. Solvency,. Solvency. ์กฐ์‚ฌ๋ณด๊ณ ์„œ 2009-02 Solvency II ์˜๋ฆฌ์Šคํฌํ‰๊ฐ€๋ชจํ˜•๋ฐ ์ธก์ •๋ฐฉ๋ฒ•์—ฐ๊ตฌ ๋ณดํ—˜์—ฐ๊ตฌ์› BIS Solvency 2007 2009 4 (RBC). 2010. Solvency. Solvency,. Solvency. โ… . ์„œ๋ก  2 ์กฐ์‚ฌ๋ณด๊ณ ์„œ 2009-02 โ…ก. Solvency II ์˜์ฒด๊ณ„๋ฐํŠน์ง• 1. ์ถ”์ง„๋ฐฐ๊ฒฝ๋ฐ์ผ์ • ์š”์•ฝ 3 2. Solvency โ…ก ์˜ 3 ์ธต์ฒด๊ณ„

More information

Microsoft Word - 130722 Hanwha Daily_New.doc

Microsoft Word - 130722 Hanwha Daily_New.doc Eagle eye-6529ํ˜ธ Today s Issue ์ด์Šˆ ๋ฏธ๊ตญ IT ๊ธฐ์—… ์‹ค์  ๋ถ€์ง„ ๋ฌด๋””์Šค, ๋ฏธ๊ตญ ์‹ ์šฉ๋“ฑ๊ธ‰ ์ „๋ง ์ƒํ–ฅ ์™ธ๊ตญ์ธ ๋ฐ ๊ธฐ๊ด€ ์ˆ˜๊ธ‰ ๊ฐœ์„ ์„ธ ๋ฏธํก ๊ธฐ์ƒ๋„ NOT GOOD GOOD NOT BAD ํˆฌ์ž์ปจ์„คํŒ…ํŒŒํŠธใ…ฃ 2013. 7. 22 Market Data ์ฃผ์š” ์ง€ํ‘œ ์ข…๊ฐ€ ์ฃผ์š” ์ง€ํ‘œ ์ข…๊ฐ€ KOSPI 1,871.41 (-0.22%) DOW 15,543.74

More information

< ๋ชฉ ์ฐจ > <์š” ์•ฝ> i I. ๊ฑฐ์‹œ๊ฒฝ์ œ ์ „๋ง 1 1. ๊ตญ๋‚ด ๊ฒฝ๊ธฐ์ƒํ™ฉ ํŒ๋‹จ 1 2. ๋Œ€์™ธ์—ฌ๊ฑด ์ „๋ง 3 3. 2016๋…„ ๊ตญ๋‚ด๊ฒฝ์ œ ์ „๋ง 10 II. ์ฃผ์š” ์‚ฐ์—… ์ „๋ง 2 1. 2015๋…„ ๋™ํ–ฅ 2 2. 2016๋…„ ์ „๋ง 24

< ๋ชฉ ์ฐจ > <์š” ์•ฝ> i I. ๊ฑฐ์‹œ๊ฒฝ์ œ ์ „๋ง 1 1. ๊ตญ๋‚ด ๊ฒฝ๊ธฐ์ƒํ™ฉ ํŒ๋‹จ 1 2. ๋Œ€์™ธ์—ฌ๊ฑด ์ „๋ง 3 3. 2016๋…„ ๊ตญ๋‚ด๊ฒฝ์ œ ์ „๋ง 10 II. ์ฃผ์š” ์‚ฐ์—… ์ „๋ง 2 1. 2015๋…„ ๋™ํ–ฅ 2 2. 2016๋…„ ์ „๋ง 24 ๋ฌธ์˜ ๋ณด๋„์ž๋ฃŒ ๋ณธ ์ž๋ฃŒ๋Š” 11์›” 26์ผ(๋ชฉ) ์กฐ๊ฐ„ ๋ถ€ํ„ฐ ๋ณด๋„ํ•ด ์ฃผ์‹œ๊ธฐ ๋ฐ”๋ž๋‹ˆ๋‹ค. ๊ฐ•๋‘์šฉ(๊ฑฐ์‹œ) (044) 287-3205 ์กฐ์ฒ (์‚ฐ์—…) (044) 287-3046 2016๋…„ ๊ฒฝ์ œ ์‚ฐ์—… ์ „๋ง 2015. 11. 25 < ๋ชฉ ์ฐจ > i I. ๊ฑฐ์‹œ๊ฒฝ์ œ ์ „๋ง 1 1. ๊ตญ๋‚ด ๊ฒฝ๊ธฐ์ƒํ™ฉ ํŒ๋‹จ 1 2. ๋Œ€์™ธ์—ฌ๊ฑด ์ „๋ง 3 3. 2016๋…„ ๊ตญ๋‚ด๊ฒฝ์ œ ์ „๋ง 10 II. ์ฃผ์š”

More information

*1601์ธํ”Œ๋ ˆ์ด์…˜01-๋ชฉ์ฐจ1~10

*1601์ธํ”Œ๋ ˆ์ด์…˜01-๋ชฉ์ฐจ1~10 ISSN 87-381 16. 1 16. 1 i i i i i 5.9 5.8 8.1 8. -.1 17.5.7 5. 6..6 9.5 9.1 7.6 8.. 3.1.9 18. 17.3 -.9. 19.5 16.9 16.9. 3.5 3.9 3.9 3.8 i 1 15 3..6.6.5.5.5 1....3.. 1.6.1 -.1...5.. -.1.1.1.1.7.9

More information

โ…  ๊ธ€๋กœ๋ฒŒ ํ™˜๊ฒฝ ์ ๊ฒ€ 1. ๋ฏธ๊ตญ์˜ ๊ฒฝ์ œ ์นจ์ฒด์— ๋Œ€ํ•œ ๋ถˆ์•ˆ๊ฐ ๋”์šฑ ํ™•์‚ฐ 1) ์ œ๋ฐ˜ ์ง€ํ‘œ์˜ ์•…ํ™” ํ™•์ธ - ๋‹ค์†Œ๊ฐ„์˜ ์‹œ์ฐจ๋Š” ์žˆ๊ฒ ์ง€๋งŒ ์ง€๋‚œํ•ด 4๋ถ„๊ธฐ ๋ฏธ๊ตญ์˜ GDP์„ฑ์žฅ์œจ์€.4%๋กœ ์˜ˆ์ƒ์น˜(1.2%)๋ฅผ ํ•˜ํšŒํ–ˆ๊ณ  1์›”๋ง์—๋Š” 28๋…„ IMF์˜ ๋ฏธ๊ตญ๊ฒฝ์ œ์„ฑ์žฅ์œจ ์ „๋ง์น˜๊ฐ€ ๋‘๋‹ฌ์—ฌ๋งŒ์— ํ•˜ํ–ฅ๋จ.

โ…  ๊ธ€๋กœ๋ฒŒ ํ™˜๊ฒฝ ์ ๊ฒ€ 1. ๋ฏธ๊ตญ์˜ ๊ฒฝ์ œ ์นจ์ฒด์— ๋Œ€ํ•œ ๋ถˆ์•ˆ๊ฐ ๋”์šฑ ํ™•์‚ฐ 1) ์ œ๋ฐ˜ ์ง€ํ‘œ์˜ ์•…ํ™” ํ™•์ธ - ๋‹ค์†Œ๊ฐ„์˜ ์‹œ์ฐจ๋Š” ์žˆ๊ฒ ์ง€๋งŒ ์ง€๋‚œํ•ด 4๋ถ„๊ธฐ ๋ฏธ๊ตญ์˜ GDP์„ฑ์žฅ์œจ์€.4%๋กœ ์˜ˆ์ƒ์น˜(1.2%)๋ฅผ ํ•˜ํšŒํ–ˆ๊ณ  1์›”๋ง์—๋Š” 28๋…„ IMF์˜ ๋ฏธ๊ตญ๊ฒฝ์ œ์„ฑ์žฅ์œจ ์ „๋ง์น˜๊ฐ€ ๋‘๋‹ฌ์—ฌ๋งŒ์— ํ•˜ํ–ฅ๋จ. ๋ฒ ์–ด๋งˆ์ผ“๋ž ๋ฆฌ๋Š” ๊ฐ€๋Šฅํ•  ์ „๋ง ๊ณจ๋“ ๋ธŒ๋ฆฟ์ง€์ฆ๊ถŒ ๋ฆฌ์„œ์น˜ํŒ€ โ…  ๊ธ€๋กœ๋ฒŒ ํ™˜๊ฒฝ ์ ๊ฒ€ 1. ๋ฏธ๊ตญ์˜ ๊ฒฝ์ œ ์นจ์ฒด์— ๋Œ€ํ•œ ๋ถˆ์•ˆ๊ฐ ๋”์šฑ ํ™•์‚ฐ 1) ์ œ๋ฐ˜ ์ง€ํ‘œ์˜ ์•…ํ™” ํ™•์ธ - ๋‹ค์†Œ๊ฐ„์˜ ์‹œ์ฐจ๋Š” ์žˆ๊ฒ ์ง€๋งŒ ์ง€๋‚œํ•ด 4๋ถ„๊ธฐ ๋ฏธ๊ตญ์˜ GDP์„ฑ์žฅ์œจ์€.4%๋กœ ์˜ˆ์ƒ์น˜(1.2%)๋ฅผ ํ•˜ํšŒํ–ˆ๊ณ  1์›”๋ง์—๋Š” 28๋…„ IMF์˜ ๋ฏธ๊ตญ๊ฒฝ์ œ์„ฑ์žฅ์œจ ์ „๋ง์น˜๊ฐ€ ๋‘๋‹ฌ์—ฌ๋งŒ์— ํ•˜ํ–ฅ๋จ. ์ผ๊ฐ์—์„œ๋Š” 2-3๋ถ„๊ธฐ์— (-)์„ฑ ์žฅ, ์˜ฌํ•ด

More information

์ธ์ ์ž์›๊ฐœ๋ฐœ์ •์ฑ… ํ˜‘๋ ฅ๋ง <์ฐจ ๋ก€> I. ๋ฌธ์ œ ์ œ๊ธฐ II. ์šฐ๋ฆฌ๋‚˜๋ผ ๊ณ ๋ นํ™”์˜ ํŠน์ง•๊ณผ ๋ฌธ์ œ์  1. ๊ณ ๋ นํ™”์˜ ํŠน์ง• (1) ์šฐ๋ฆฌ๋‚˜๋ผ ์ธ๊ตฌ 10๋ช…์ค‘ 1๋ช…์ด ๋…ธ์ธ (2) ๋†์–ด์ดŒ ์ง€์—ญ์˜ ์ดˆ๊ณ ๋ น์‚ฌํšŒํ™” (3) ๋…ธ์ธ์ธ๊ตฌ์˜ ์œ ๋…„์ธ๊ตฌ ์ถ”์›”(2016๋…„) (4) ์ƒ์‚ฐ๊ฐ€๋Šฅ์ธ๊ตฌ 7.3๋ช…์ด ๋…ธ์ธ 1

์ธ์ ์ž์›๊ฐœ๋ฐœ์ •์ฑ… ํ˜‘๋ ฅ๋ง <์ฐจ ๋ก€> I. ๋ฌธ์ œ ์ œ๊ธฐ II. ์šฐ๋ฆฌ๋‚˜๋ผ ๊ณ ๋ นํ™”์˜ ํŠน์ง•๊ณผ ๋ฌธ์ œ์  1. ๊ณ ๋ นํ™”์˜ ํŠน์ง• (1) ์šฐ๋ฆฌ๋‚˜๋ผ ์ธ๊ตฌ 10๋ช…์ค‘ 1๋ช…์ด ๋…ธ์ธ (2) ๋†์–ด์ดŒ ์ง€์—ญ์˜ ์ดˆ๊ณ ๋ น์‚ฌํšŒํ™” (3) ๋…ธ์ธ์ธ๊ตฌ์˜ ์œ ๋…„์ธ๊ตฌ ์ถ”์›”(2016๋…„) (4) ์ƒ์‚ฐ๊ฐ€๋Šฅ์ธ๊ตฌ 7.3๋ช…์ด ๋…ธ์ธ 1 ์ธ์ ์ž์›์ •์ฑ…ํ˜‘๋ ฅ๋ง ์ •๋ก€ํ† ๋ก  ์ œ1์ฃผ์ œ ๊ณ ๋ นํ™”์‚ฌํšŒ์˜ ๊ณ ๋ น์ž ์ธ์ ์ž์›๊ด€๋ฆฌ ๋ฐœํ‘œ์ž: ์œ ์„ฑ๊ธฐ(ํ•œ๊ตญ์ธ์‚ฌ๊ด€๋ฆฌํ•™ํšŒ ๋ถ€ํšŒ์žฅ) ์ธ์ ์ž์›๊ฐœ๋ฐœ์ •์ฑ… ํ˜‘๋ ฅ๋ง I. ๋ฌธ์ œ ์ œ๊ธฐ II. ์šฐ๋ฆฌ๋‚˜๋ผ ๊ณ ๋ นํ™”์˜ ํŠน์ง•๊ณผ ๋ฌธ์ œ์  1. ๊ณ ๋ นํ™”์˜ ํŠน์ง• (1) ์šฐ๋ฆฌ๋‚˜๋ผ ์ธ๊ตฌ 10๋ช…์ค‘ 1๋ช…์ด ๋…ธ์ธ (2) ๋†์–ด์ดŒ ์ง€์—ญ์˜ ์ดˆ๊ณ ๋ น์‚ฌํšŒํ™” (3) ๋…ธ์ธ์ธ๊ตฌ์˜ ์œ ๋…„์ธ๊ตฌ ์ถ”์›”(2016๋…„) (4) ์ƒ์‚ฐ๊ฐ€๋Šฅ์ธ๊ตฌ

More information

26(3D)-17.fm

26(3D)-17.fm 26ฦ’ 3D ร 2006 5ล“ pp. ~ ยช y w qp yw k d Predictive Equation of Dynamic Modulus for Hot Mix Asphalt with Granite Aggregates yรกยฝx ร Lee, Kwan-HoรKim, Hyun-OรJang, Min-Seok Abstract The presented work provided

More information

ๆญฏํ‘œ์ง€.PDF

ๆญฏํ‘œ์ง€.PDF 2000 - - 2001 3 LG i < > 1 2 1,000, 90 10 50-70% TV,, PC, 1/3,, 90,, 90, TV,, TV,, TV,,,,,,,,,, 10 20 2000 55 ( ), ( ) ii iii 2,,,,,, < > < > I 1 1 1 2 3 II 5 1 5 2 10 III 16 1 16 2 22 3 24 4 27 5 31 IV

More information

๋ฏธ๊ตญ์˜ ๊ฒฝ์ œ๋™ํ–ฅ๊ณผ ์–‘์ ์™„ํ™” ํ…Œ์ดํผ๋ง ์ „๋ง (์š”์•ฝ๋ฌธ) ์œค๋•๋ฃก(๋Œ€์™ธ๊ฒฝ์ œ์ •์ฑ…์—ฐ๊ตฌ์›, ์„ ์ž„์—ฐ๊ตฌ์œ„์›), ์ด๋™์€(๋Œ€์™ธ๊ฒฝ์ œ์ •์ฑ…์—ฐ๊ตฌ์›, ๊ตญ์ œ๊ฑฐ์‹œํŒ€์žฅ) 1. ๋ฏธ๊ตญ ๊ฒฝ์ œ ๋™ํ–ฅ ๋ฏธ๊ตญ ๊ฒฝ์ œ์ง€ํ‘œ๋“ค์€ ์ผ๋ถ€ ๊ต๋ž€์š”์ธ์ด ์—†๋Š” ๊ฒƒ์€ ์•„๋‹ˆ์ง€๋งŒ ์•ˆ์ •์ ์ธ ํšŒ๋ณต์„ธ๋ฅผ ์‹œํ˜„ํ•˜๊ณ  ์žˆ ์Œ. - ์‹ค์—…๋ฅ ์˜ ํ•˜๋ฝ์„ธ๊ฐ€ ์ง€์†๋˜๊ณ , ์ฃผํƒ๊ฐ€๊ฒฉ์€ ์ƒ์Šน์„ธ๋ฅผ ์œ ์ง€ํ•˜๊ณ  ์žˆ์–ด์„œ ๊ฒฝ์ œ์„ฑ์žฅ ์ถ”์„ธ๊ฐ€ ์ด ์–ด์งˆ ๊ฒƒ์„ ์‹œ์‚ฌ - ๊ฐ€๊ณ„๋ถ€์ฑ„๊ฐ€

More information

Dc_F&OWeek

Dc_F&OWeek Derivatives Spot 28. 1. 24 ์ „ ๊ท  (2)22-744 gyun.jun@samsung.com ์„ฑ์ˆ˜์—ฐ (2)22-7823 sooyeon.sung@samsung.com www.samsungfn.com Futures & Options Weekly ELS์— ๋Œ€ํ•œ ์˜คํ•ดํ’€๊ธฐ ELS์— ๋Œ€ํ•œ ๋ถ€์ •ํ™•ํ•œ ์ดํ•ด์—์„œ ๋ฒ—์–ด๋‚˜์•ผ ์ฃผ์‹์‹œ์žฅ์˜ ํญ๋ฝ ์™€์ค‘์— ELS์˜ knock-in(์›๊ธˆ๋ณด์žฅ๊ตฌ๊ฐ„

More information

Microsoft Word - Macro120103.doc

Microsoft Word - Macro120103.doc Weekly Focus ๊ตญ์ œ์œ ๊ฐ€, ์ถ”๊ฐ€ ์ƒ์Šน ๊ฐ€๋Šฅ์„ฑ์— ๋Œ€ํ•œ ์ ๊ฒ€ Economist ์ฑ„ํ˜„๊ธฐ (769-3002) coolheuy@deri.co.kr Summary 2011๋…„ ๊ตญ์ œ์œ ๊ฐ€๋Š” ์—ฐํ‰๊ท  ๊ธฐ์ค€์œผ๋กœ ์ง€๋‚œ 2010๋…„์— ๊ธฐ๋กํ–ˆ๋˜ ๋ฐฐ๋Ÿด๋‹น 79๋‹ฌ๋Ÿฌ ๋Œ€๋น„ ์•ฝ 25% ์ƒ ์Šนํ•œ ๋ฐฐ๋Ÿด๋‹น 95๋‹ฌ๋Ÿฌ ์ˆ˜์ค€์„ ๊ธฐ๋กํ•จ์— ๋”ฐ๋ผ 2๋…„ ์—ฐ์† level-up๋˜๋Š” ํ๋ฆ„์„ ์‹œํ˜„ํ•จ. 2012๋…„์—

More information

<C8A3C4A1B9CEBCF6C3E2C0CEC5A5BAA3C0CCC5CDB0B3BCD22831323037292E687770>

<C8A3C4A1B9CEBCF6C3E2C0CEC5A5BAA3C0CCC5CDB0B3BCD22831323037292E687770> ์ค‘์†Œ๊ธฐ์—…์ง„ํฅ๊ณต๋‹จ (www.sbc.or.kr) ๋ณด๋„์ž๋ฃŒ ๋ฒ ํŠธ๋‚จ ํ˜ธ์น˜๋ฏผ ไธญ ไผ ์ˆ˜์ถœ์ธํ๋ฒ ์ดํ„ฐ ๊ฐœ์†Œ 7์ผ์— ๊ฐœ์†Œํ•˜์—ฌ ๋™๋‚จ์•„ ์ˆ˜์ถœ์ „์ง„๊ธฐ์ง€ ์—ญํ•  ์ˆ˜ํ–‰ 12๊ฐœ์—…์ฒด ์ž…์ฃผ, ํ•ด์™ธ์ถœ์žฅ ์ค‘์†Œ๊ธฐ์—…์— ๋น„์ฆˆ๋‹ˆ์Šค ํŽธ์˜ ์ œ๊ณต ์ค‘์†Œ๊ธฐ์—…๋“ค ์ด ๊ธฐ์กด ์˜ ์ˆ˜์ถœ์‹œ ์žฅ์ธ ๋ฏธ ๊ตญ, ์ผ๋ณธ ๋“ฑ ์—์„œ ์‹œ ์„  ์„ ์˜ฎ ๊ฒจ ์‹œ ์žฅ ์ž ์žฌ๋ ฅ์ด ๋ฌดํ•œํ•œ ๋™๋‚จ์•„ ๋“ฑ ์„ ์ง‘ ์ค‘ ๊ณต๋žต ํ•˜๋Š” ๋…ธ ๋ ฅ ์ด ์ „๊ฐœ๋˜ ๊ณ  ์žˆ๋Š” ๊ฐ€์šด๋ฐ

More information

i - ii - iii - iv - 1 - 2 - 3 - 4 - 5 - 6 - 7 - ฯ† ฮป 8 - 9 - 10 - ฮด โธ ฮด max max max 11 -../data/1 data_directory ---- 1 top output_file_head(fohead) ---- 2 4 number_of_channel ---- 3 VV VH HV HH output_channel_head

More information

KSKSKSKS SKSKSKS KSKSKS SKSKS KSKS SKS KS KS C 3004 KS C

KSKSKSKS SKSKSKS KSKSKS SKSKS KSKS SKS KS KS C 3004 KS C KSKSKSKS SKSKSKS KSKSKS SKSKS KSKS SKS KS KS C 3004 KS C 3004 2002 2002 12 27 ICS 2906020 2904020 KS Testing methods for rubber or plastic insulated wires and cables ( ) KS B 5202 KS B 5203 KS B 5206 KS

More information

- 2 -

- 2 - 2014 ๋…„์œต ๋ณตํ•ฉ๊ธฐ์ˆ ๊ฐœ๋ฐœ์‚ฌ์—… ( ์œต ๋ณตํ•ฉ๊ณผ์ œ ) ์ œ์•ˆ์š”์ฒญ์„œ ๋ชฉ์ฐจ - 1 - - 2 - - 3 - - 4 - - 5 - - 6 - - 7 - - 8 - - 9 - - 10 - - 11 - - 12 - - 13 - - 14 - - 15 - - 16 - - 17 - ฮฉ - 18 - - 19 - - 20 - ๊ธฐ๊ณ„์†Œ์žฌ -001-21 - ๊ธฐ๊ณ„์†Œ์žฌ -002-22 - ๊ธฐ๊ณ„์†Œ์žฌ

More information

์ „ํ†ต ์†Œ์‹ธ์›€์‚ฐ์—… ์œก์„ฑ์„ ์œ„ํ•œ ๋„์„ธ๊ฐ๋ฉด ํƒ€๋‹น์„ฑ ์—ฐ๊ตฌ 2015. 7. 30. ์—ฐ๊ตฌ์ฑ…์ž„ : ๊น€ํ•„ํ—Œ ํ•œ๊ตญ์ง€๋ฐฉ์„ธ์—ฐ๊ตฌ์› ์—ฐ๊ตฌ์œ„์› ๊ณต๋™์—ฐ๊ตฌ : ์ด์ƒํ›ˆ ํ•œ๊ตญ์ง€๋ฐฉ์„ธ์—ฐ๊ตฌ์› ์—ฐ๊ตฌ์œ„์› ์—ฐ ๊ตฌ ์ง„ : ์‚ฌ๋ช…์ฒ  ํ•œ๊ตญ์ง€๋ฐฉ์„ธ์—ฐ๊ตฌ์› ์—ฐ๊ตฌ์› โ… . ์„œ๋ก  m ๊ฒฝ๋ถ๋„๋Š” ์†Œ์‹ธ์›€์ด ์ง€์—ญ๊ฒฝ์ œ์— ๋ฏธ์น˜๋Š” ๊ธ์ •์  ์˜ํ–ฅ์„ ๋ฐ˜์˜ํ•˜์—ฌ ๋ ˆ์ €์„ธ๋ฅผ ๊ฐ ๋ฉดํ•ด ์ฃผ๊ณ  ์žˆ์Œ. ๊ฒฝ๋ฅœ, ๊ฒฝ์ •, ๊ฒฝ๋งˆ ๋ฐ ์ „ํ†ต ์†Œ์‹ธ์›€๊ฒฝ๊ธฐ์—

More information

<C3D6C1BE5FBFACB1B8BAB8B0EDBCAD323031342D30335FC1B6B5BFC3B65FBFECB8AEB0E6C1A6C0C7BFAAB5BFBCBA5F6B303430312E687770>

<C3D6C1BE5FBFACB1B8BAB8B0EDBCAD323031342D30335FC1B6B5BFC3B65FBFECB8AEB0E6C1A6C0C7BFAAB5BFBCBA5F6B303430312E687770> ์—ฐ๊ตฌ๋ณด๊ณ ์„œ 2014-03 ์šฐ๋ฆฌ ๊ฒฝ์ œ์˜ ์—ญ๋™์„ฑ: ์ผ๋ณธ๊ณผ์˜ ๋น„๊ต๋ฅผ ์ค‘์‹ฌ์œผ๋กœ ์กฐ๋™์ฒ  ํŽธ ๋ฐœ๊ฐ„์‚ฌ ์šฐ๋ฆฌ ๊ฒฝ์ œ๋Š” ์ผ๋ณธ๊ฒฝ์ œ๋ฅผ ๋‹ฎ์•„ ๊ฐ€๋Š”๊ฐ€? ์ด๋Š” ํ•™๊ณ„๋ฟ ์•„๋‹ˆ๋ผ ๊ธฐ์—…, ์–ธ๋ก , ์‹ฌ์ง€์–ด ์ผ๋ฐ˜์ธ๋“ค๋„ ์ž์ฃผ ํ•˜๋Š” ์งˆ๋ฌธ์ด๋‹ค. ์ด ์งˆ๋ฌธ์—๋Š” ์ผ๋ณธ์ด ๊ฑฐํ’ˆ๊ฒฝ์ œ์˜ ๋ถ• ๊ดด๋ฅผ ๊ฒช์œผ๋ฉด์„œ ์žฅ๊ธฐ์นจ์ฒด์— ์ง„์ž…ํ•˜๋˜ 1990๋…„๋Œ€์˜ ์ƒํ™ฉ์„ ๋– ์˜ฌ๋ฆฌ๋ฉด์„œ, ์šฐ๋ฆฌ ๊ฒฝ์ œ์—๋„ ๋ถˆ๊ธธํ•œ ์žฅ๋ž˜๊ฐ€ ์˜ˆ๊ณ ๋˜์–ด ์žˆ๋Š” ๊ฒƒ์€ ์•„๋‹Œ๊ฐ€ ํ•˜๋Š”

More information

8, Aug, 08, 008

8, Aug, 08, 008 8, Aug, 08, 008 8, Aug, 08, 008 3 8, Aug, 08, 008 4 8, Aug, 08, 008 5 8, Aug, 08, 008 6 8, Aug, 08, 008 7 8, Aug, 08, 008 8 8, Aug, 08, 008 9 8, Aug, 08, 008 0 8, Aug, 08, 008 8, Aug, 08, 008 0.00% KTB

More information

Microsoft Word - 2016081214212698K_01_01.docx

Microsoft Word - 2016081214212698K_01_01.docx Global Market Economics 216. 8. 12 Week Ahead: 8์›” ์…‹์งธ ์ฃผ ํ•˜๋ฐ˜๊ธฐ ๋ฏธ๊ตญ ์ธํ”Œ๋ ˆ์ด์…˜ ์ •์ƒํ™” ๊ธฐ๋Œ€ Global Macro Team Analysts ํ—ˆ์ง„์šฑ (Global/US) james_huh@samsung.com 2 22 7796 ์ด์Šนํ›ˆ (China/Korea) sh213.lee@samsung.com 2 22 7836 ๊น€์ง€์€

More information

ๆญฏ25์ฐจ.PPT

ๆญฏ25์ฐจ.PPT 25 1. 1 25 GDP 6.3% GDP : 2/4 8.0% (99 1/4 4.6%) : 5 21.8% (1/4 12.5%) : 5 89.6 % (1/4-45.3%) 2/4 75% (BSI) 120 2 25 98 99 (, %) 1/4 4 5 1/4 4 5-6.2-11.4-11.1-7.3 12.5 17.1 21.8-4.5-10.3-11.8-7.1 14.1

More information

212๋…„ ํ•˜๋ฐ˜๊ธฐ ๊ธˆ๋ฆฌ์ „๋ง ๋ฐ ์ฑ„๊ถŒํˆฌ์ž์ „๋žต ๊ทธ๋ฆผ 1 ์ฃผ์š”๊ตญ ๊ตญ์ฑ„1๋…„ ๊ธˆ๋ฆฌ ์ถ”์ด (%) 8 7 6 5 4 3 2 1 Spain Italy Korea Malaysia China Australia US UK Germany Japan 27 28 29 21 211 212 ์ž๋ฃŒ:

212๋…„ ํ•˜๋ฐ˜๊ธฐ ๊ธˆ๋ฆฌ์ „๋ง ๋ฐ ์ฑ„๊ถŒํˆฌ์ž์ „๋žต ๊ทธ๋ฆผ 1 ์ฃผ์š”๊ตญ ๊ตญ์ฑ„1๋…„ ๊ธˆ๋ฆฌ ์ถ”์ด (%) 8 7 6 5 4 3 2 1 Spain Italy Korea Malaysia China Australia US UK Germany Japan 27 28 29 21 211 212 ์ž๋ฃŒ: ์กฐ์‚ฌ ์—ฐ๊ตฌ 212๋…„ ํ•˜๋ฐ˜๊ธฐ ๊ธˆ๋ฆฌ์ „๋ง ๋ฐ ์ฑ„๊ถŒํˆฌ์ž์ „๋žต ์‹ ๋™์ค€ / ๋™๋ถ€์ฆ๊ถŒ ๋ฆฌ์„œ์น˜์„ผํ„ฐ ํˆฌ์ž์ „๋žต๋ณธ๋ถ€์žฅ โ… . 212๋…„ ํ•˜๋ฐ˜๊ธฐ ๊ธˆ๋ฆฌ์ „๋ง โ…ก. ์œ ๋กœ์กด ํ•ด๋ฒ•๊ณผ ๊ตญ๋‚ด์™ธ ํŽ€๋”๋ฉ˜ํ„ธ ์—ฌ๊ฑด ์ ๊ฒ€ โ…ข. ํ†ตํ™”์ •์ฑ…๊ณผ ๋ฌผ๊ฐ€, ์™ธ๊ตญ์ธ์˜ ์ž๊ธˆ์ดํƒˆ ์šฐ๋ ค โ…ฃ. ๊ธ€๋กœ๋ฒŒ๊ฒฝ์ œ์™€ Re-decoupling Story โ… . 212๋…„ ํ•˜๋ฐ˜๊ธฐ ๊ธˆ๋ฆฌ์ „๋ง ์•ˆ์ „์ž์‚ฐ ์„ ํ˜ธ๋กœ ๊ธ€๋กœ๋ฒŒ ๊ตญ์ฑ„๊ธˆ๋ฆฌ๋Š” ์‚ฌ์ƒ ์ตœ์ €์น˜๊นŒ์ง€ ํ•˜๋ฝ

More information